Skeptics and scientists both attempt to employ the same scientific method. The difference is that scientists focus on hypotheses that are promising, whereas skeptics focus on hypotheses that are popular.
This can have ramifications. For example, I've noticed that a lot of people with first-hand scientific experience (inc. my dad, for what it's worth) don't have a lot of respect for Karl Popper's attempts to distinguish science from non-science and identify a distinct scientific process. They tend to see science as working in a more intuitive, less formalised, fashion.
That's for two reasons. Firstly, scientists are mainly surrounded by promising hypotheses generated by other competent scientists. Skeptics aren't - we deal with the daftest ideas that human ingenuity can devise. And secondly, if a scientist comes across a non-promising and unsupported idea, they are "allowed" to discard it out of hand. Skeptics aren't, because then the cranks and quacks start whining that we're not treating them fairly.
Going the other way, I know many skeptics* are disturbed by the climate science community's complete PR failure in the case of "Climategate" and other alleged scandals. "Why did they not see this coming?" we ask ourselves. "Have they never dealt with people like this before?"
No, they haven't. They're scientists, not skeptics. They spend all their time working through complex climate models and looking at atmospheric data. They've never argued with a creationist, or tried to talk sense into a Holocaust denier. They're not equipped for this kind of debate.
Scientists and skeptics are natural complements, and there is a lot of overlap between the two groups. Skeptics have always taken scientists seriously; hopefully in the future this will go both ways.
* By this I do not mean climate change pseudoskeptics. I should mention that I don't have the knowledge (yet) to evaluate the claims of modern climate science, so I can't comment on whether it's correct. But what I do know is that most of the main claims made by climate change deniers are ludicrously simplistic and/or just plain wrong.
Read the full post
Tuesday, May 11, 2010
British law: Bloody inconsistent
It's been barely a month since the brilliant response by the Court of Appeals in the BCA vs Simon Singh case. Sadly, they can't all be gems. Paul Chambers has just been convicted.
In case anyone hasn't come across this case, Mr Chambers made a comment on his Twitter feed expressing frustration at the closure of his local airport:
Pretty normal for Twitter, you'd think, and frankly rather mild for the Internet as a whole. Sadly, the Crown Prosecution Service didn't agree. When they got wind of this comment, they decided to prosecute him for sending a message on a public network that was "grossly offensive or of an indecent, obscene or menacing character".
Quite apart from the WTF that we have such a fuzzy law on our books, there is no evidence that Mr Chambers intended to menace anyone, or that anyone in fact felt menaced. This is apparently a case of the CPS deciding that Mr Chambers must be guilty of something, and then hunting up an obscure law to fit. (We actually have a law specifically dealing with bomb threats, but that would have required the CPS to make a more solid case so they (ab)used an obscenity law instead.)
Now, as an average netizen, I'd normally respond with a comment like:
But given the CPS's apparent inability to recognise hyperbolic comments, I wouldn't dream of saying any such thing.
Read the full post
In case anyone hasn't come across this case, Mr Chambers made a comment on his Twitter feed expressing frustration at the closure of his local airport:
"Crap! Robin Hood Airport is closed. You've got a week and a bit to get your shit together, otherwise I’m blowing the airport sky high!"
Pretty normal for Twitter, you'd think, and frankly rather mild for the Internet as a whole. Sadly, the Crown Prosecution Service didn't agree. When they got wind of this comment, they decided to prosecute him for sending a message on a public network that was "grossly offensive or of an indecent, obscene or menacing character".
Quite apart from the WTF that we have such a fuzzy law on our books, there is no evidence that Mr Chambers intended to menace anyone, or that anyone in fact felt menaced. This is apparently a case of the CPS deciding that Mr Chambers must be guilty of something, and then hunting up an obscure law to fit. (We actually have a law specifically dealing with bomb threats, but that would have required the CPS to make a more solid case so they (ab)used an obscenity law instead.)
Now, as an average netizen, I'd normally respond with a comment like:
"These vindictive fascist little jobsworths will be first against the wall when the revolution comes."
But given the CPS's apparent inability to recognise hyperbolic comments, I wouldn't dream of saying any such thing.
Read the full post
Sunday, April 25, 2010
Status report
Just so you're not surprised by continued silence on my part... the last post was by way of a stress outlet. I'm still not done with exams, I'm just past the point where I can bring myself to revise very hard.
Exams end in 1 week. The traditional post-exams hangover will take a bit longer. It usually takes me a couple extra weeks to reacquaint myself with the concept of a social life (and in particular with those parts of my social circle that like to drink large amounts of CH3CH2OH). So expect to hear from me within the month.
Read the full post
Exams end in 1 week. The traditional post-exams hangover will take a bit longer. It usually takes me a couple extra weeks to reacquaint myself with the concept of a social life (and in particular with those parts of my social circle that like to drink large amounts of CH3CH2OH). So expect to hear from me within the month.
Read the full post
Gazing towards Olympus
I don't normally write fiction because... well... I suck. No way of sugar-coating it, my writing stinks.
That said, I would rather it didn't always stink. So in recent months, in between episodes of screaming exam panic, I've been casting around for stories to experiment with. This one popped more or less fully-formed into my head. I think it was sparked by this Penny Arcade comic about the game God Of War III.
I actually have no idea what GoW3 is about. My version is probably cooler though.
Now, I know what you're thinking: a bird that talks? But the plain truth is, you can't live thousands of years without picking up a few things.
I'm not actually sure how long it was. Certainly in the early days I was just an eagle like any other. Every morning I'd glide over to that mountain, just as the dawn struck it, ready for another breakfast of liver-and-onions. Without the onions, obviously. And with a lot more screaming. But it was a simple life, with a noticeable absence of clocks and calendars. Could have been millennia for all I know.
I think it was some time during the Roman occupation that he started speaking to me. Not surprising really - you chain a guy to a sodding great big rock in the boiling sun for a few hundred years, he'll wind up talking to everything from stones to flying pigs. That's dehydration for you. Eagles is small fry by comparison.
But it was nice, you know? Even when I was just another dumb beast, it was soothing. I'll let you in on a little secret: even birds of prey can get a bit squeamish at times. Cute little lambs gambolling in the green meadows, claws out, whoosh... you'd have to have a pretty stony heart not to feel slightly sorry for them. Not to feel a touch of existential angst on occasion. It was reassuring to know that at least one of my victims didn't hold it against me.
So he spoke to me, and for a century or ten that was all there was to it. There's precedent, you know: prisoners in towers befriending the sparrows, and all that. Befriending the eagle that rips out your viscera every morning is a bit of a stretch, but he always was soft-hearted. That's how he got into this mess in the first place.
And, then as now, soft-heartedness can have some... startling effects. Some time just after they started building churches, it clicked. I began to understand what these funny sounds he made were in aid of. I'm still quite proud of that accomplishment. I mean, even Champollion had the Rosetta stone to help him, you know what I mean? And already spoke a language, for that matter. I was operating from a bit more of a standing start.
Of course this didn't really mean much for fifty years or so, the eagle throat not being noticeably designed for speaking back. But I think he realised I could follow him. He started trying to turn his ramblings into a dialogue, inviting me to nod my head or shake my tail or whatever. He was more than half-mad with hunger and thirst by this point, but we eventually worked out a sort of talon-tapping routine that could get the message across.
(Incidentally this proved to be a useful investment in fine motor skill. After all, how do you think I'm writing this? My typing speed isn't great, but as long as I can get access to a keyboard... Let's just say that some web-cafes shouldn't leave their skylight unlocked, haha.)
Actually he mostly talked about you lot. Humans I mean. He really couldn't get over whether he'd done the right thing or not, giving you fire. He would often ask me to go out and report back to him on how you were doing, what works of art you were producing, what buildings you were constructing. I remember spending a week one time trying to explain the Sistine Chapel to him. I mean, come on, the Sistine Chapel? In glorified Morse code? But he got the general impression at least.
Come to think of it, that was maybe not so tactful, showing him how the world had moved on. Zeus had traded his thunderbolts in for a new throne, the rest of the Pantheon had reinvented themselves as angels and saints and the like, and yet here he was still chained to a mountain. They'd forgotten him. And all the art of mankind, that he had helped bring into being, depicted him as a serpent or monster. You lot aren't exactly much for gratitude.
And I suppose that was the next big development. I started to see his point. He really did want the best for you, you know. Most of the gods spent their time hanging round mountaintops; he was down in the valleys with you. Depressingly keen on helping people, three thousand years before Robert Baden-Powell.
He didn't deserve his punishment. I don't know if Zeus had got up on the wrong side of the cloud that day or what, but damn. Chained to a rock with yours truly performing impromptu surgery on a daily basis. That's gotta sting.
I started eating less of his liver. Some days I'd just make a scratch, give him a bit of an appendectomy scar in case anyone dropped by to check. I brought him berries, fruit, water. Took a while to work out what he was able to stomach (I swear I didn't know about the peanut allergy!) but eventually he started to get healthier. You don't exactly shake off millennia of torment overnight, but he began to seem a little more like his old self.
Sadder, of course. Maybe wiser. Certainly more angry. But hey, at least he'd stopped screaming. My eardrums were endlessly grateful.
And that brings us up to the present day. Or at least up to the day - a week ago now - when he asked me to help him. When I agreed to take the final step.
It wasn't easy. People look at you funny if you, an eagle, walk into the local library and check out a book on lock-picking. And some of the other stuff he was asking for... well, normally if someone's packing that kind of hardware, any wise bird avoids them on pain of taxidermy. But there are ways. As I said before, you don't live for thousands of years without picking up a thing or three.
So here we are. Or here I am, anyway. His chains are empty, his rock is bare. I can see the indentation he wore in it from lying there so long. And I can remember the look in his eye as his hands caressed the rifle's stock. I remember how he gazed towards Olympus.
I guess I'm still waiting for the other shoe to fall. What happens if Zeus comes looking for me, afterwards? Or... what happens if he doesn't? Either way, it's going to be one hell of a light-show.
I never told Prometheus about the Crusades, or the Holocaust, or napalm, or the bomb. I couldn't stand to see the look on his face if he'd found out what that fire he carried had been used for. If he'd known the cost of giving fire to man. Eating his liver would have been nothing by comparison.
But I think I know what he would have said. The true fire, the true spark of warmth, is the one in the mind and the heart and the eyes. The physical hearth and forge, the technology, is just a conduit for the greatness in man. I understand this because I felt it too. I felt it when my beak fumbled with the lockpicks, when I dropped the heavy gun by his side.
He gave fire to man. I gave it back.
I hope we were right.
Read the full post
That said, I would rather it didn't always stink. So in recent months, in between episodes of screaming exam panic, I've been casting around for stories to experiment with. This one popped more or less fully-formed into my head. I think it was sparked by this Penny Arcade comic about the game God Of War III.
I actually have no idea what GoW3 is about. My version is probably cooler though.
Now, I know what you're thinking: a bird that talks? But the plain truth is, you can't live thousands of years without picking up a few things.
I'm not actually sure how long it was. Certainly in the early days I was just an eagle like any other. Every morning I'd glide over to that mountain, just as the dawn struck it, ready for another breakfast of liver-and-onions. Without the onions, obviously. And with a lot more screaming. But it was a simple life, with a noticeable absence of clocks and calendars. Could have been millennia for all I know.
I think it was some time during the Roman occupation that he started speaking to me. Not surprising really - you chain a guy to a sodding great big rock in the boiling sun for a few hundred years, he'll wind up talking to everything from stones to flying pigs. That's dehydration for you. Eagles is small fry by comparison.
But it was nice, you know? Even when I was just another dumb beast, it was soothing. I'll let you in on a little secret: even birds of prey can get a bit squeamish at times. Cute little lambs gambolling in the green meadows, claws out, whoosh... you'd have to have a pretty stony heart not to feel slightly sorry for them. Not to feel a touch of existential angst on occasion. It was reassuring to know that at least one of my victims didn't hold it against me.
So he spoke to me, and for a century or ten that was all there was to it. There's precedent, you know: prisoners in towers befriending the sparrows, and all that. Befriending the eagle that rips out your viscera every morning is a bit of a stretch, but he always was soft-hearted. That's how he got into this mess in the first place.
And, then as now, soft-heartedness can have some... startling effects. Some time just after they started building churches, it clicked. I began to understand what these funny sounds he made were in aid of. I'm still quite proud of that accomplishment. I mean, even Champollion had the Rosetta stone to help him, you know what I mean? And already spoke a language, for that matter. I was operating from a bit more of a standing start.
Of course this didn't really mean much for fifty years or so, the eagle throat not being noticeably designed for speaking back. But I think he realised I could follow him. He started trying to turn his ramblings into a dialogue, inviting me to nod my head or shake my tail or whatever. He was more than half-mad with hunger and thirst by this point, but we eventually worked out a sort of talon-tapping routine that could get the message across.
(Incidentally this proved to be a useful investment in fine motor skill. After all, how do you think I'm writing this? My typing speed isn't great, but as long as I can get access to a keyboard... Let's just say that some web-cafes shouldn't leave their skylight unlocked, haha.)
Actually he mostly talked about you lot. Humans I mean. He really couldn't get over whether he'd done the right thing or not, giving you fire. He would often ask me to go out and report back to him on how you were doing, what works of art you were producing, what buildings you were constructing. I remember spending a week one time trying to explain the Sistine Chapel to him. I mean, come on, the Sistine Chapel? In glorified Morse code? But he got the general impression at least.
Come to think of it, that was maybe not so tactful, showing him how the world had moved on. Zeus had traded his thunderbolts in for a new throne, the rest of the Pantheon had reinvented themselves as angels and saints and the like, and yet here he was still chained to a mountain. They'd forgotten him. And all the art of mankind, that he had helped bring into being, depicted him as a serpent or monster. You lot aren't exactly much for gratitude.
And I suppose that was the next big development. I started to see his point. He really did want the best for you, you know. Most of the gods spent their time hanging round mountaintops; he was down in the valleys with you. Depressingly keen on helping people, three thousand years before Robert Baden-Powell.
He didn't deserve his punishment. I don't know if Zeus had got up on the wrong side of the cloud that day or what, but damn. Chained to a rock with yours truly performing impromptu surgery on a daily basis. That's gotta sting.
I started eating less of his liver. Some days I'd just make a scratch, give him a bit of an appendectomy scar in case anyone dropped by to check. I brought him berries, fruit, water. Took a while to work out what he was able to stomach (I swear I didn't know about the peanut allergy!) but eventually he started to get healthier. You don't exactly shake off millennia of torment overnight, but he began to seem a little more like his old self.
Sadder, of course. Maybe wiser. Certainly more angry. But hey, at least he'd stopped screaming. My eardrums were endlessly grateful.
And that brings us up to the present day. Or at least up to the day - a week ago now - when he asked me to help him. When I agreed to take the final step.
It wasn't easy. People look at you funny if you, an eagle, walk into the local library and check out a book on lock-picking. And some of the other stuff he was asking for... well, normally if someone's packing that kind of hardware, any wise bird avoids them on pain of taxidermy. But there are ways. As I said before, you don't live for thousands of years without picking up a thing or three.
So here we are. Or here I am, anyway. His chains are empty, his rock is bare. I can see the indentation he wore in it from lying there so long. And I can remember the look in his eye as his hands caressed the rifle's stock. I remember how he gazed towards Olympus.
I guess I'm still waiting for the other shoe to fall. What happens if Zeus comes looking for me, afterwards? Or... what happens if he doesn't? Either way, it's going to be one hell of a light-show.
I never told Prometheus about the Crusades, or the Holocaust, or napalm, or the bomb. I couldn't stand to see the look on his face if he'd found out what that fire he carried had been used for. If he'd known the cost of giving fire to man. Eating his liver would have been nothing by comparison.
But I think I know what he would have said. The true fire, the true spark of warmth, is the one in the mind and the heart and the eyes. The physical hearth and forge, the technology, is just a conduit for the greatness in man. I understand this because I felt it too. I felt it when my beak fumbled with the lockpicks, when I dropped the heavy gun by his side.
He gave fire to man. I gave it back.
I hope we were right.
Read the full post
Saturday, April 03, 2010
Fluffy nonsense
I hate Business Economics, indeed. With a passion. I hate it so much because the cello part is the worst cello part ever written in the history of
Wait, wrong rant.
I'm currently studying for actuarial exam CT7* (Business Economics). Essentially, the purpose of Business Economics is to determine when a sale will occur and at what price. I've hated this subject from the get-go, but it's taken me some time to really articulate what I dislike about it...
What is Business Economics?
To understand Business Economics, the first thing you need to know is: price is determined by the demand for a product and the supply of that product. When the demand at a given price equals the supply at that price, the market is in equilibrium and all produce will be sold. Markets tend towards equilibrium.
The supply function is determined by the seller's return on investment at various prices. A seller will pick the price that maximises their RoI. You can draw a "supply curve" showing the amount that will be produced at various prices. This will be a fairly straight upwards-sloping line.
The demand function is determined by the buyer's utility function. This is a subjective measure of how much value a buyer gets from a product, denominated in currency (e.g. £). Buyers will seek to maximise their total consumer excess: their total utility minus the products' purchase price. You can draw a "demand curve" to show how the amount bought varies at different prices, which will be a fairly straight downwards-sloping line.
To really understand Business Economics, the second thing you need to know is: all of the above is counterfactual bollocks. Markets don't really work like that under any circumstances. You can't draw demand or supply curves, and they wouldn't look like their descriptions if you could.
Theology meets practicality
These two different views of BE are both presented in the same textbook. Chapters 4-7 and 9-12 describe everything about a business in terms of demand and supply. Chapter 8, however, discusses the marketing mix.
The marketing mix is a widely-referenced checklist of things to consider when selling a product. It lists the following components of a successful sale:
1) Product
2) Price
3) Placement (e.g. location, visibility)
4) Promotion (e.g. advertising)
The traditional demand/supply model handles the first two of these fine, but is completely blind to the last two. The only way to account for e.g. the effect of advertising is to say "well, that just means that the person's utility function is higher for this product than it would otherwise have been".
But this just highlights the problem with the demand/supply model. It's not answering the question of "what determines price". It's just restating it in a more sciencey fashion. It is theology rather than science: no more meaningful than answering the question "what created everything" with "God did it".
In short, it sucks.
What went wrong?
I'm not sure why it sucks so much. My best guess is that the people who came up with this theory of economics had "engineering envy". In engineering, you start with various general principles (the laws of science) and reason your way to specific applications of those principles (bridges, buildings, microcircuits, etc).
I would conjecture that the early economists rather liked this idea, so they came up with some vague principles (the "laws" of demand and supply) and tried to reason from them to specific conclusions. But because those broad principles were not reality-based, the effect is similar to a Christian trying to reason from the existence of the Trinity to the existence of an earthquake in Haiti. It's an exercise in applied fuzziness, goalpost-moving, equivocation and redefinition of terms that would shame a Jesuit***.
This entire approach sucks.
First Principles
I'm still working on my own Grand Unified Theory of Economics, and probably will be for some time. But I already know where to start looking.
Look at the big picture again, and start to zoom in. Zoom in on a single continent (Asia). Zoom in on a single country (India). Zoom in on a single region (Uttar Pradesh). Zoom in on a town (Agra, home of the Taj Mahal). Zoom in on a little shop selling tables made of marble, and inlaid with semiprecious stones. Zoom in on the room where a tourist and a salesman are sipping tea together.
That tourist is me. That salesman, I am fairly sure, overcharged me by an order of magnitude for what was, I must admit, a rather lovely souvenir for my parents.
When an economic model can explain what happened in that room, in concepts more meaningful than the data they abstract...
When an economic model can explain what happened in that room, in a way that helps me to identify and avoid such situations in future...
When an economic model can explain what happened in that room, in enough detail that I can see how to return the "favour"...
...then I will consider it worth learning for exams.
Rant over.
----------------------------
* "CT" stands for "Core Technical". To become a fully fledged Fellow of the Institute of Actuaries** I need to have completed:
9 x Core Technical exams
3 x Core Applied exams
2 x Specialist Technical exams (from a choice of 9)
1 x Specialist Applied exam (from a choice of 6)
1 x Partridge in a Pear Tree.
** In Britain, for historical reasons, we have both an Institute of Actuaries and a Fellowship of Actuaries (collectively known as the Profession). Most Cambridge grads join the IoA. This is so that, when we've claimed our congregational MA and achieved the rank of Fellow of the IoA, we can put the letters MA FIA after our name and pretend to be Sicilian dons.
*** Yes, I know that the Jesuits were actually quite science-minded for their time. Their reputation for inflicting "sophisticated" theological arguments on the uninitiated is probably still valid.
Read the full post
Wait, wrong rant.
I'm currently studying for actuarial exam CT7* (Business Economics). Essentially, the purpose of Business Economics is to determine when a sale will occur and at what price. I've hated this subject from the get-go, but it's taken me some time to really articulate what I dislike about it...
What is Business Economics?
To understand Business Economics, the first thing you need to know is: price is determined by the demand for a product and the supply of that product. When the demand at a given price equals the supply at that price, the market is in equilibrium and all produce will be sold. Markets tend towards equilibrium.
The supply function is determined by the seller's return on investment at various prices. A seller will pick the price that maximises their RoI. You can draw a "supply curve" showing the amount that will be produced at various prices. This will be a fairly straight upwards-sloping line.
The demand function is determined by the buyer's utility function. This is a subjective measure of how much value a buyer gets from a product, denominated in currency (e.g. £). Buyers will seek to maximise their total consumer excess: their total utility minus the products' purchase price. You can draw a "demand curve" to show how the amount bought varies at different prices, which will be a fairly straight downwards-sloping line.
To really understand Business Economics, the second thing you need to know is: all of the above is counterfactual bollocks. Markets don't really work like that under any circumstances. You can't draw demand or supply curves, and they wouldn't look like their descriptions if you could.
Theology meets practicality
These two different views of BE are both presented in the same textbook. Chapters 4-7 and 9-12 describe everything about a business in terms of demand and supply. Chapter 8, however, discusses the marketing mix.
The marketing mix is a widely-referenced checklist of things to consider when selling a product. It lists the following components of a successful sale:
1) Product
2) Price
3) Placement (e.g. location, visibility)
4) Promotion (e.g. advertising)
The traditional demand/supply model handles the first two of these fine, but is completely blind to the last two. The only way to account for e.g. the effect of advertising is to say "well, that just means that the person's utility function is higher for this product than it would otherwise have been".
But this just highlights the problem with the demand/supply model. It's not answering the question of "what determines price". It's just restating it in a more sciencey fashion. It is theology rather than science: no more meaningful than answering the question "what created everything" with "God did it".
In short, it sucks.
What went wrong?
I'm not sure why it sucks so much. My best guess is that the people who came up with this theory of economics had "engineering envy". In engineering, you start with various general principles (the laws of science) and reason your way to specific applications of those principles (bridges, buildings, microcircuits, etc).
I would conjecture that the early economists rather liked this idea, so they came up with some vague principles (the "laws" of demand and supply) and tried to reason from them to specific conclusions. But because those broad principles were not reality-based, the effect is similar to a Christian trying to reason from the existence of the Trinity to the existence of an earthquake in Haiti. It's an exercise in applied fuzziness, goalpost-moving, equivocation and redefinition of terms that would shame a Jesuit***.
This entire approach sucks.
First Principles
I'm still working on my own Grand Unified Theory of Economics, and probably will be for some time. But I already know where to start looking.
Look at the big picture again, and start to zoom in. Zoom in on a single continent (Asia). Zoom in on a single country (India). Zoom in on a single region (Uttar Pradesh). Zoom in on a town (Agra, home of the Taj Mahal). Zoom in on a little shop selling tables made of marble, and inlaid with semiprecious stones. Zoom in on the room where a tourist and a salesman are sipping tea together.
That tourist is me. That salesman, I am fairly sure, overcharged me by an order of magnitude for what was, I must admit, a rather lovely souvenir for my parents.
When an economic model can explain what happened in that room, in concepts more meaningful than the data they abstract...
When an economic model can explain what happened in that room, in a way that helps me to identify and avoid such situations in future...
When an economic model can explain what happened in that room, in enough detail that I can see how to return the "favour"...
...then I will consider it worth learning for exams.
Rant over.
----------------------------
* "CT" stands for "Core Technical". To become a fully fledged Fellow of the Institute of Actuaries** I need to have completed:
9 x Core Technical exams
3 x Core Applied exams
2 x Specialist Technical exams (from a choice of 9)
1 x Specialist Applied exam (from a choice of 6)
1 x Partridge in a Pear Tree.
** In Britain, for historical reasons, we have both an Institute of Actuaries and a Fellowship of Actuaries (collectively known as the Profession). Most Cambridge grads join the IoA. This is so that, when we've claimed our congregational MA and achieved the rank of Fellow of the IoA, we can put the letters MA FIA after our name and pretend to be Sicilian dons.
*** Yes, I know that the Jesuits were actually quite science-minded for their time. Their reputation for inflicting "sophisticated" theological arguments on the uninitiated is probably still valid.
Read the full post
Friday, April 02, 2010
Contagion effects
One interesting topic of financial research in recent years has been the concept of "contagion". This is the effect whereby (for example) financial troubles at one bank can lead to a "domino effect", potentially bringing down the entire damn economy...
(This sort of thing is very important from an actuarial perspective - it goes back to my discussion of stochastic asset models. It's common to try to invest in a low-risk fund when a policyholder nears retirement; however, in a contagion-ridden market, there may be no such thing as a low-risk fund!)
I have three comments on this:
1) At the same time I was in Cambridge last week, they were holding a conference on this very subject. I've been reading through some of the papers, and they're rather cool - I might summarise some on this blog later.
2) There seems to be a strong link to principles of ecosystem collapse (see for example this excellent paper from PLoS CompBio). I'm surprised there's not more cross-research going on. (Or maybe there is and I just haven't found it?)
3) Some kind of contagion principle seems to apply to the use of the term "in respect of". You can measure how heavily-regulated someone's industry is just by seeing how often they employ this phrase. After some consideration, I have decided that it can almost always be replaced with the word "for". This, of course, is vastly shorter, thus saving wear and tear on keyboards.
OK, so that third point didn't really relate to the other two, but it's been bugging me for a while. As someone with a love of the English language and its intricacies, I really hate it when people try to fake linguistic sophistication by use of stock phrases and pompous legal jargon. It's like I'm a fan of automobile engineering and they're the blokes from "Pimp My Ride".
Read the full post
(This sort of thing is very important from an actuarial perspective - it goes back to my discussion of stochastic asset models. It's common to try to invest in a low-risk fund when a policyholder nears retirement; however, in a contagion-ridden market, there may be no such thing as a low-risk fund!)
I have three comments on this:
1) At the same time I was in Cambridge last week, they were holding a conference on this very subject. I've been reading through some of the papers, and they're rather cool - I might summarise some on this blog later.
2) There seems to be a strong link to principles of ecosystem collapse (see for example this excellent paper from PLoS CompBio). I'm surprised there's not more cross-research going on. (Or maybe there is and I just haven't found it?)
3) Some kind of contagion principle seems to apply to the use of the term "in respect of". You can measure how heavily-regulated someone's industry is just by seeing how often they employ this phrase. After some consideration, I have decided that it can almost always be replaced with the word "for". This, of course, is vastly shorter, thus saving wear and tear on keyboards.
OK, so that third point didn't really relate to the other two, but it's been bugging me for a while. As someone with a love of the English language and its intricacies, I really hate it when people try to fake linguistic sophistication by use of stock phrases and pompous legal jargon. It's like I'm a fan of automobile engineering and they're the blokes from "Pimp My Ride".
Read the full post
Interesting turn of phrase
I'm having a lazy day.
This is probably not a good idea - actuarial exams start in under 3 weeks. However, I'm still recovering from last weekend, when I received my congregational MA at Cambridge and got absolutely wasted on College port* with some old friends. I vaguely recall dancing on tables. And I had a busy week too, so what the hey.
In particular, I'm taking the time to read up on a few blog archives, mostly related to skepticism in the UK. Of particular interest recently has been the Appeal Court decision in the trial of science writer Simon Singh. The (very senior) judges running the show apparently gave the legal version of two raised fingers to the British Chiropractic Organisation. A most welcome verdict.
(I actually walk past the BCA's head office on my way into work each day. I'm considering dropping some Sense About Science literature through their door in case they feel like signing up. Worth a try.)
One interesting thought for the day was in this pro-CAM comment on the Adventures In Nonsense blog. FYI, the blog author is the guy who sent out 500 complaints to the Advertising Standards Authority about chiropractors who claimed to treat infant colic.
The comment uses an interesting phrase: "open-minded skeptic". I am actually impressed that a pro-alternative-medicine commenter would use this phrase; many of them seem to think it's an oxymoron. But let's examine what it means to say a person is an "open-minded skeptic".
"Open-minded" = is willing to examine any new claim
"Sceptic" (with a C) = demands good evidence for any new claim before accepting it
"Skeptic" (with a K) = gets pissed off if he/she receives no good evidence and yet finds people still parroting the same sodding claim
On this basis, I would say that most of the folks on Simon Singh's side in this battle of public opinion are indeed "open-minded skeptics". I'm happy to quietly support them; I only wish I could do more.
------------------
* If you ever visit Christ's bar, do not buy the College red or white wine, it gives you a splitting hangover. The port, however, kicks ass but leaves neurons standing. There's also a cafe called Taffy's just out the College's side gate where you can buy a fry-up cure for what ails ya the following morning.
Read the full post
This is probably not a good idea - actuarial exams start in under 3 weeks. However, I'm still recovering from last weekend, when I received my congregational MA at Cambridge and got absolutely wasted on College port* with some old friends. I vaguely recall dancing on tables. And I had a busy week too, so what the hey.
In particular, I'm taking the time to read up on a few blog archives, mostly related to skepticism in the UK. Of particular interest recently has been the Appeal Court decision in the trial of science writer Simon Singh. The (very senior) judges running the show apparently gave the legal version of two raised fingers to the British Chiropractic Organisation. A most welcome verdict.
(I actually walk past the BCA's head office on my way into work each day. I'm considering dropping some Sense About Science literature through their door in case they feel like signing up. Worth a try.)
One interesting thought for the day was in this pro-CAM comment on the Adventures In Nonsense blog. FYI, the blog author is the guy who sent out 500 complaints to the Advertising Standards Authority about chiropractors who claimed to treat infant colic.
The comment uses an interesting phrase: "open-minded skeptic". I am actually impressed that a pro-alternative-medicine commenter would use this phrase; many of them seem to think it's an oxymoron. But let's examine what it means to say a person is an "open-minded skeptic".
"Open-minded" = is willing to examine any new claim
"Sceptic" (with a C) = demands good evidence for any new claim before accepting it
"Skeptic" (with a K) = gets pissed off if he/she receives no good evidence and yet finds people still parroting the same sodding claim
On this basis, I would say that most of the folks on Simon Singh's side in this battle of public opinion are indeed "open-minded skeptics". I'm happy to quietly support them; I only wish I could do more.
------------------
* If you ever visit Christ's bar, do not buy the College red or white wine, it gives you a splitting hangover. The port, however, kicks ass but leaves neurons standing. There's also a cafe called Taffy's just out the College's side gate where you can buy a fry-up cure for what ails ya the following morning.
Read the full post
Wednesday, March 24, 2010
Cheers, Mate
Thanks to this guy for cutting short my evening-long search for a way to rip CDs to MP3.
He's right, it is ridiculously hard on Linux. My understanding is that that's because That Can Get You Sued, and Canonical is probably big enough to be a tasty target for the MP3 patent holders.
Normally I would use the ogg format, but I just got a cheap mp3 player thingy and none of the cheap ones seem to handle ogg. I did check.
Read the full post
He's right, it is ridiculously hard on Linux. My understanding is that that's because That Can Get You Sued, and Canonical is probably big enough to be a tasty target for the MP3 patent holders.
Normally I would use the ogg format, but I just got a cheap mp3 player thingy and none of the cheap ones seem to handle ogg. I did check.
Read the full post
Wednesday, March 17, 2010
How to do Stochastic Calculus
Context
I have recently had to learn a mathematical approach called "stochastic calculus" as part of my actuarial exams. As misery loves company, I've decided to share the details with you...
This post is a short tutorial in stochastic calculus, with an emphasis on getting to the point where we can solve practical problems and skipping as much foundational material as I can get away with.
What is stochastic calculus?
Stochastic calculus is a useful tool in financial maths. In normal calculus, you might take a function and find its derivatives (gradient, curvature, etc) as time changes. Or you might take a differential equation (an equation relating a function to its derivatives) and use it to figure out what the corresponding function looks like.
"Stochastic" just means probability-related. In stochastic calculus, you take a random variable and find its derivatives, or take a differential equation and find the random variable it represents.
An example. Let's say we're looking at the size (nt) of a population of lemurs. We reckon that the rate of population change is directly proportional to the size of the population. In mathematical notation:
dnt/dt = g.nt
where g is a constant growth factor (or shrinkage factor if negative). If we solved this with differential calculus we'd find that:
nt = n0.exp(g.t)
But wait! Real populations don't work like that. They behave far more randomly, which can lead to very different results. In particular, you'll notice that, no matter how negative the growth rate is, the lemurs can never go extinct. Sadly this is not true of real animals.
How do we represent this uncertainty? One way is to add a "noise" term to the calculation. So:
dNt/dt = g.Nt + v.Wt.Nt
where v is a constant volatility factor and Wt is a continuously-changing random variable - a "stochastic process". You'll note that we're now using Nt rather than nt, to show that population size is probabilistic rather than predestined.
Brownian motion
Which stochastic process should we use for our noise term? This would depend very much on the situation - our team of biologists would need to research how species' populations change in practice. However, a very popular choice is based on "Brownian motion". It's typical to try to reframe any stochastic problem in terms of this process.
Brownian motion is named after the phenomenon of pollen grains bouncing around in water (they're small enough to be affected by the impact of individual water molecules). It is a "random walk" process, with several nice properties:
- it is distributed as N(0,t2) (e.g the standard deviation at time t is t0.5)
- it has independent increments, so you can model B2t as Bt + B't
- dBt.dBt = dt (please take it on trust that this is important, we'll see how it's relevant later)
In the case of our population equation, we might set Wt such that Wt.dt = dBt. In other words, Bt is in some sense the integral between times 0 and t of Wt. This allows us to rephrase our equation as:
dNt = g.Nt.dNt + v.Nt.Wt.dt
= g.Nt.dNt + v.Nt.dBt
Interlude: What the hell is that?
I keep using terms like dBt and dNt to mean increments of random variables. Intuitively, this doesn't make a heck of a lot of sense. Random variables, even time-dependent ones, don't have fixed values - that's kinda the point - so how the blazes can they have well-defined increments?
The answer is: the increment is also a random variable, just an infinitesimally small one (in the same way that dt is an infinitesimally small deterministic variable). When we say (for example) I = ∫0t 1.ds, the answer is clearly: I=t. When we say I = ∫0t 1.dBs, the answer is equally obvious: I = Bt, the sum of the random increments.
More complex expressions like ∫0t Bs.dBs are quite hard to understand intuitively. Multiplying a random variable by an infinitesimal random variable and taking an infinite sum? Wtf? Basically, the purpose of stochastic calculus is to allow us to transform messy situations like this back into simple ones like the two in the preceding paragraph.
The Ito formula
So anyway, we now know what the stochastic differential equation looks like. How do we solve it? We use an equation called the Ito formula to figure out what it should look like, and then we forcibly bring the two forms together.
The Ito formula says that if:
dXt = u.dt + v.dBt
(u, v real numbers)
and:
Yt = f(t, Xt) for some function f(t,x)
then:
dYt = ∂f/∂t.dt + ∂f/∂x.dXt + 1/2.∂2f/∂2x.dXt.dXt
using e.g. ∂f/∂x as a shorthand for ∂f/∂x(t, Xt)
Usually it makes sense to take Xt = Bt.
A simple example. Let's say we're evaluating It = ∫ Bt.dBt. We can rephrase this as: dIt = Bt.dBt
The variable It will end up depending on time and on the effects of the random variable, so we can write: It = f(t, Bt). Therefore, by Ito's formula:
dIt = ∂f/∂t.dt + ∂f/∂x.dBt + 1/2.∂2f/∂2x.dBt.dBt
= (∂f/∂t + 1/2.∂2f/∂2x).dt + ∂f/∂x.dBt because dBt.dBt = dt (see the section on Brownian motion above).
Since we know: dIt = 0.dt + Bt.dBt
we can say that: ∂f/∂x = x
and: ∂f/∂t = -1/2.∂2f/∂2x
From the first of these partial derivatives, we have:
f = x2/2 + g(t)
We can now rewrite the second equality as: ∂f/∂t = -1/2.1
giving: f = -t/2 + h(x)
So overall: f = x2/2 - t/2
Therefore (drumroll), we have: It = f(t,Bt) = Bt2/2 - t/2
That population equation
Now we've got the tools in place, let's solve that population equation. First assume that there is some function f(t,x) such that Nt = f(t,Bt)
We know that:
dNt = g.Nt.dNt + v.Nt.dBt
and that:
dNt = (∂f/∂t - 1/2.∂2f/∂2x).dt + ∂f/∂x.dBt
Therefore:
∂f/∂x = v.f
f = exp(v.x).p(t)
(I'm using p(t) and q(x) as my dummy functions rather than g(t) and h(x), to avoid confusion between g and g(t))
And:
∂f/∂t + 1/2.∂2f/∂2x = g.f
∂f/∂t + 1/2.v2.f = g.f
∂f/∂t = (g - v2/2).f
f = exp((g - v2/2).t).q(x)
So overall we can infer that:
f = C.exp(v.x + (g-v2/2).t) for constant C
In other words:
Nt = N0.exp(v.Bt + (g-v2/2).t)
QED. This is similar, but not identical to what we'd expect from looking at the traditional differential equation.
How we can use this
Now, just having an equation is not terribly useful. But once we've got it we can start asking interesting questions such as "what is the probability that the population of lemurs will have shrunk at time t?"
P(Nt < N0) = P(N0.exp(v.Bt + (g-v2/2).t) < N0)
= P(exp(v.Bt + (g-v2/2).t) < 1)
= P(v.Bt + (g-v2/2).t < 0)
= P(Bt < (v2/2-g).t/v)
Since Bt ~ N(0,t), this can be worked out using a set of tables. For example if g = 5%, v = 2%, and t = 5, then:
P(N5 < N0) = P(B5 < (0.0004/2 - 0.05)*5/0.02)
= P(Z < (0.0004/2 - 0.05)*5/0.02/5^0.5)
= P(Z < (0.0004/2 - 0.05)*5/0.02/5^0.5)
= P(X < -5.5678)
Past a certain point the population of lemurs is likely to go extinct (although this is not modelled well by the equation - Nt can never reach 0). If the population was declining, and we could figure out values of g and v, we could determine the probability of extinction. We could then allocate our animal conservation money appropriately.
Applying these principles to pension fund capital is left as an exercise to the reader.
Why is that not the full story?
There are several things I've left out here, two mathematical points and two practical points.
Firstly, Wt is not a "real" distribution. If you actually try to work out what dBt/dt looks like, you find that it is infinite at infinitely many points. That is just plain weird.
Fortunately, there is a mathematical hold-all concept - the "extended distribution" that sweeps all that under the carpet. And since it all works out OK, the specifics aren't really a concern, any more than you need to know Real Analysis to understand differential calculus.
Or are they a concern? Point #2 is that there are actually several ways to define ∫...dBt, which give mutually exclusive results. Explaining the differences is beyond the scope of this post. Suffice to say that the one we've been using - the Ito integral - is mathematically nice.
Practically speaking, what we've looked at here is only half the battle. As responsible scientists, we'd want to produce sensible values for g and v, and in general confirm that the population is behaving as the equation predicts. This calibration is a nightmarish task that I'm not even going to begin to discuss here.
And finally there's the question of whether our general approach - all founded, you'll recall, on Brownian motion - is remotely sensible. There has been a lot of muttering recently pointing out that extreme events are far more likely than the normal distribution would suggest. It is very hard to produce Brownian processes that can emulate this behaviour, and nigh-on impossible to calibrate. There is a strong suggestion that a broader family called Levy processes might be more plausible, but they are also a bugger to calibrate.
This sort of thing won't be solved any time soon.
References
I learned SC from the textbook by Bernt Oksendal (5th edition). In retrospect this was probably a mistake, as he goes very deeply into the mathematical fundamentals before covering anything remotely practical. If you can get past the first 3 chapters without falling asleep then it's not a bad book, although not terribly readable (even by maths book standards...).
Read the full post
I have recently had to learn a mathematical approach called "stochastic calculus" as part of my actuarial exams. As misery loves company, I've decided to share the details with you...
This post is a short tutorial in stochastic calculus, with an emphasis on getting to the point where we can solve practical problems and skipping as much foundational material as I can get away with.
What is stochastic calculus?
Stochastic calculus is a useful tool in financial maths. In normal calculus, you might take a function and find its derivatives (gradient, curvature, etc) as time changes. Or you might take a differential equation (an equation relating a function to its derivatives) and use it to figure out what the corresponding function looks like.
"Stochastic" just means probability-related. In stochastic calculus, you take a random variable and find its derivatives, or take a differential equation and find the random variable it represents.
An example. Let's say we're looking at the size (nt) of a population of lemurs. We reckon that the rate of population change is directly proportional to the size of the population. In mathematical notation:
dnt/dt = g.nt
where g is a constant growth factor (or shrinkage factor if negative). If we solved this with differential calculus we'd find that:
nt = n0.exp(g.t)
But wait! Real populations don't work like that. They behave far more randomly, which can lead to very different results. In particular, you'll notice that, no matter how negative the growth rate is, the lemurs can never go extinct. Sadly this is not true of real animals.
How do we represent this uncertainty? One way is to add a "noise" term to the calculation. So:
dNt/dt = g.Nt + v.Wt.Nt
where v is a constant volatility factor and Wt is a continuously-changing random variable - a "stochastic process". You'll note that we're now using Nt rather than nt, to show that population size is probabilistic rather than predestined.
Brownian motion
Which stochastic process should we use for our noise term? This would depend very much on the situation - our team of biologists would need to research how species' populations change in practice. However, a very popular choice is based on "Brownian motion". It's typical to try to reframe any stochastic problem in terms of this process.
Brownian motion is named after the phenomenon of pollen grains bouncing around in water (they're small enough to be affected by the impact of individual water molecules). It is a "random walk" process, with several nice properties:
- it is distributed as N(0,t2) (e.g the standard deviation at time t is t0.5)
- it has independent increments, so you can model B2t as Bt + B't
- dBt.dBt = dt (please take it on trust that this is important, we'll see how it's relevant later)
In the case of our population equation, we might set Wt such that Wt.dt = dBt. In other words, Bt is in some sense the integral between times 0 and t of Wt. This allows us to rephrase our equation as:
dNt = g.Nt.dNt + v.Nt.Wt.dt
= g.Nt.dNt + v.Nt.dBt
Interlude: What the hell is that?
I keep using terms like dBt and dNt to mean increments of random variables. Intuitively, this doesn't make a heck of a lot of sense. Random variables, even time-dependent ones, don't have fixed values - that's kinda the point - so how the blazes can they have well-defined increments?
The answer is: the increment is also a random variable, just an infinitesimally small one (in the same way that dt is an infinitesimally small deterministic variable). When we say (for example) I = ∫0t 1.ds, the answer is clearly: I=t. When we say I = ∫0t 1.dBs, the answer is equally obvious: I = Bt, the sum of the random increments.
More complex expressions like ∫0t Bs.dBs are quite hard to understand intuitively. Multiplying a random variable by an infinitesimal random variable and taking an infinite sum? Wtf? Basically, the purpose of stochastic calculus is to allow us to transform messy situations like this back into simple ones like the two in the preceding paragraph.
The Ito formula
So anyway, we now know what the stochastic differential equation looks like. How do we solve it? We use an equation called the Ito formula to figure out what it should look like, and then we forcibly bring the two forms together.
The Ito formula says that if:
dXt = u.dt + v.dBt
(u, v real numbers)
and:
Yt = f(t, Xt) for some function f(t,x)
then:
dYt = ∂f/∂t.dt + ∂f/∂x.dXt + 1/2.∂2f/∂2x.dXt.dXt
using e.g. ∂f/∂x as a shorthand for ∂f/∂x(t, Xt)
Usually it makes sense to take Xt = Bt.
A simple example. Let's say we're evaluating It = ∫ Bt.dBt. We can rephrase this as: dIt = Bt.dBt
The variable It will end up depending on time and on the effects of the random variable, so we can write: It = f(t, Bt). Therefore, by Ito's formula:
dIt = ∂f/∂t.dt + ∂f/∂x.dBt + 1/2.∂2f/∂2x.dBt.dBt
= (∂f/∂t + 1/2.∂2f/∂2x).dt + ∂f/∂x.dBt because dBt.dBt = dt (see the section on Brownian motion above).
Since we know: dIt = 0.dt + Bt.dBt
we can say that: ∂f/∂x = x
and: ∂f/∂t = -1/2.∂2f/∂2x
From the first of these partial derivatives, we have:
f = x2/2 + g(t)
We can now rewrite the second equality as: ∂f/∂t = -1/2.1
giving: f = -t/2 + h(x)
So overall: f = x2/2 - t/2
Therefore (drumroll), we have: It = f(t,Bt) = Bt2/2 - t/2
That population equation
Now we've got the tools in place, let's solve that population equation. First assume that there is some function f(t,x) such that Nt = f(t,Bt)
We know that:
dNt = g.Nt.dNt + v.Nt.dBt
and that:
dNt = (∂f/∂t - 1/2.∂2f/∂2x).dt + ∂f/∂x.dBt
Therefore:
∂f/∂x = v.f
f = exp(v.x).p(t)
(I'm using p(t) and q(x) as my dummy functions rather than g(t) and h(x), to avoid confusion between g and g(t))
And:
∂f/∂t + 1/2.∂2f/∂2x = g.f
∂f/∂t + 1/2.v2.f = g.f
∂f/∂t = (g - v2/2).f
f = exp((g - v2/2).t).q(x)
So overall we can infer that:
f = C.exp(v.x + (g-v2/2).t) for constant C
In other words:
Nt = N0.exp(v.Bt + (g-v2/2).t)
QED. This is similar, but not identical to what we'd expect from looking at the traditional differential equation.
How we can use this
Now, just having an equation is not terribly useful. But once we've got it we can start asking interesting questions such as "what is the probability that the population of lemurs will have shrunk at time t?"
P(Nt < N0) = P(N0.exp(v.Bt + (g-v2/2).t) < N0)
= P(exp(v.Bt + (g-v2/2).t) < 1)
= P(v.Bt + (g-v2/2).t < 0)
= P(Bt < (v2/2-g).t/v)
Since Bt ~ N(0,t), this can be worked out using a set of tables. For example if g = 5%, v = 2%, and t = 5, then:
P(N5 < N0) = P(B5 < (0.0004/2 - 0.05)*5/0.02)
= P(Z < (0.0004/2 - 0.05)*5/0.02/5^0.5)
= P(Z < (0.0004/2 - 0.05)*5/0.02/5^0.5)
= P(X < -5.5678)
Past a certain point the population of lemurs is likely to go extinct (although this is not modelled well by the equation - Nt can never reach 0). If the population was declining, and we could figure out values of g and v, we could determine the probability of extinction. We could then allocate our animal conservation money appropriately.
Applying these principles to pension fund capital is left as an exercise to the reader.
Why is that not the full story?
There are several things I've left out here, two mathematical points and two practical points.
Firstly, Wt is not a "real" distribution. If you actually try to work out what dBt/dt looks like, you find that it is infinite at infinitely many points. That is just plain weird.
Fortunately, there is a mathematical hold-all concept - the "extended distribution" that sweeps all that under the carpet. And since it all works out OK, the specifics aren't really a concern, any more than you need to know Real Analysis to understand differential calculus.
Or are they a concern? Point #2 is that there are actually several ways to define ∫...dBt, which give mutually exclusive results. Explaining the differences is beyond the scope of this post. Suffice to say that the one we've been using - the Ito integral - is mathematically nice.
Practically speaking, what we've looked at here is only half the battle. As responsible scientists, we'd want to produce sensible values for g and v, and in general confirm that the population is behaving as the equation predicts. This calibration is a nightmarish task that I'm not even going to begin to discuss here.
And finally there's the question of whether our general approach - all founded, you'll recall, on Brownian motion - is remotely sensible. There has been a lot of muttering recently pointing out that extreme events are far more likely than the normal distribution would suggest. It is very hard to produce Brownian processes that can emulate this behaviour, and nigh-on impossible to calibrate. There is a strong suggestion that a broader family called Levy processes might be more plausible, but they are also a bugger to calibrate.
This sort of thing won't be solved any time soon.
References
I learned SC from the textbook by Bernt Oksendal (5th edition). In retrospect this was probably a mistake, as he goes very deeply into the mathematical fundamentals before covering anything remotely practical. If you can get past the first 3 chapters without falling asleep then it's not a bad book, although not terribly readable (even by maths book standards...).
Read the full post
Tuesday, February 02, 2010
Sucks to be us
I get a lot of emails from the Institute of Actuaries, the professional organisation of which I am a (student) member. One running theme is a rather plaintive request for people to view actuaries as more trustworthy. They write endless editorials on actuarial ethics, they arrange super-special "professionalism" training, they commit us all to worryingly restrictive codes of conduct... the list is endless.
I've always been a bit surprised by this. I don't have a pension, I don't have life insurance, I don't have a car to insure, so pretty much my first contact with the actuarial profession was when I signed up to become one. I've been wondering for some time now what the fuss is all about...
Well, it turns out that they're not wrong to fuss. I just came across the following passage in a book on scams that I'm reading:
Wait a minute. That's us he's talking about. This sucks.
Do actuaries deserve this slagging-off? Well, partly yes, partly no. (Equivocation never harmed anyone...)
Where's the money?
Let's take an example: that with-profits thing he mentioned. With-profits is superficially rather a clever idea. Let's say you're putting money into a pensions fund for your retirement. There are two main kinds of investment fund: fixed interest and unit-linked.
With a fixed interest fund you are guaranteed (say) a 3% per year rate of return. In a way this certainty is good - that money is as good as in the bank. But in a way it's bad, because you are going to get a very crappy rate of return. This is because, if the market crashes, that 2% is going to start looking like quite a large number. No fund provider wants to be caught out that way, so they will all low-ball their figures to compensate.
With a unit-linked fund, a certain amount of money is put into a pot for you. That money is used to buy shares in various companies, in the hope that their share price will go up. Usually you'll just specify some broad criteria ("I want all my money in Asian shares") and a fund manager will sort out the investment specifics. This approach generally gives a much higher rate of return - e.g. you might get 10% instead of 3%. But it is inherently risky - you might also get -50%.
As pensions actuaries, we want to get the best of both worlds. How can we do this? There are three main approaches (that I'm aware of):
1) Shuffle money from unitised to fixed-interest funds as the member approaches retirement (I mentioned this last time as a good use of SAMs)
2) Invest in an index-linked fund (see footnote*)
3) Invest in a with-profits fund
Technical interlude
A with-profits fund is where you take a fixed interest fund and strap a unitised fund to the roof. The fund provider will keep a pot of money, which it will invest in various companies etc. They will guarantee you a relatively low rate of return (3%, say) on your contribution to this pot.
Then, at the end of every year, they will take a look at how their investments have done. If the investments are doing much better than the guaranteed rate (10%, say), they will tell you "we declare a bonus of (say) 5%". The value of your investment will thus increase by 8%, with the remaining 2% being kept in the pot for a rainy day**.
The point is that, under the with-profits model, the fund provider doesn't need to plan ahead so carefully - if they overdo the bonuses this year, they can just make it up by under-doing it in future years. With-profits funds allow the provider to wing it. They don't have to be as cautious as they are with fixed-interest funds. And that reduced risk translates directly into better rates of return.
What are the characteristics of a with-profit fund? In any given year, there's a limit to how badly they can do (handy if a recession strikes just before you're due to retire). However, in the long term, they won't do as well as unit-linked funds (because they have that extra protection, which must be paid for). They're generally quite a good solution, and fill a very useful niche.
With-profiteering
Now, being smart people, you've probably noticed a gap in the above explanation: how precisely is the bonus calculated? That's an excellent question.
And the answer is very fuzzy. The provider's actuaries will look at the state of the market, look at the state of the fund, and make an educated guess as to what bonus they can afford. There is no hard-and-fast method, and no requirement to document their method. From the point of view of the paying public, the actuaries all go into a dark room and come out a week later saying "right, you're getting 8%". In the words of my generation: What The F***?
Now there are actually good reasons for some of this reticence. An example: many of these calculations use stochastic asset models (see last post) that are just too damn complicated to explain to the general public. Due to the huge pile of regulation hanging anvil-like over our heads, it can often be legally safer to shut up than explain what you're doing. Especially if there's any chance of being creatively misunderstood***.
Regardless of this, the opacity of the system - the secrecy, the lack of information, the argument from authority - definitely promotes suspicion. It's a fricken' invitation to conspiracy theory. I personally would be surprised if the bonus system had ever been abused too badly, just because there are so many people breathing down the average actuary's neck. But you can see how people could come to that conclusion, and there's no way to refute them. Skepticism FAIL.
What can the Institute of Actuaries do about this? Bugger-all, sadly. They don't have the power to compel all these companies to open up, which is really the only way to fix the situation. Hence the rather frantic "oh we're so professional and ethical" song and dance - it's the only thing the poor sods can do to stop their collective street-cred going down the pipe.
Who does have that power? The UK government (in the UK, anyway) and its regulator the Financial Services Authority. Will they act? On existing funds this is very unlikely - it would be too much of a can of worms to open. However, the general trend is towards greater transparency with financial information****. Maybe one day we'll get this mess sorted out.
And the profession will breathe a sigh of relief.
Disclaimer: I am not a subject matter expert, please take all the above with a pinch of salt.
* An index-linked fund guarantees you (say) 1% plus your central bank's base interest rate per year. This removes some of the risk that the buying power of your money will fall (e.g. due to hyperinflation). So although the amount of money you'll get when you retire is uncertain, the number of mars bars you'll be able to buy with it is closer to being fixed.
In terms of returns, index-linked funds are less good than unitised funds, but not nearly as bad as traditional fixed-interest funds. In terms of risks, the order is reversed (as you'd expect).
** There are some complexities (e.g. various bonus types, market value reductions, etc) - this is just an overview not a technical manual.
*** In the UK, the life office Equitable Life went bankrupt at least partly because A) people managed to creatively misunderstand their pensions' guarantee structure, and B) a court decided to hold them to the most expensive interpretation.
**** For example, under the new Solvency II requirements (basically saying how much money an insurer needs to put by in case everything goes wrong at once), the rather complicated models used by each company to calculate their funding requirements must be disclosed. I consider this a very positive step, not least because I want to play around with the models (and possibly blog about them!).
Read the full post
I've always been a bit surprised by this. I don't have a pension, I don't have life insurance, I don't have a car to insure, so pretty much my first contact with the actuarial profession was when I signed up to become one. I've been wondering for some time now what the fuss is all about...
Well, it turns out that they're not wrong to fuss. I just came across the following passage in a book on scams that I'm reading:
Hundreds of thousands, if not millions, have been disappointed and forced to complain about 'with-profits' bonds, endowments and pension plans. And these all come from highly respected - and really big - life insurance companies.
How do these complicated schemes work? Well, no one really knows, often not even the people who should know. It's all down to the 'experts' at the insurance companies which run these schemes. Using obscure mathematics coupled with a finger held up to the wind, they make up the numbers year on year.
Wait a minute. That's us he's talking about. This sucks.
Do actuaries deserve this slagging-off? Well, partly yes, partly no. (Equivocation never harmed anyone...)
Where's the money?
Let's take an example: that with-profits thing he mentioned. With-profits is superficially rather a clever idea. Let's say you're putting money into a pensions fund for your retirement. There are two main kinds of investment fund: fixed interest and unit-linked.
With a fixed interest fund you are guaranteed (say) a 3% per year rate of return. In a way this certainty is good - that money is as good as in the bank. But in a way it's bad, because you are going to get a very crappy rate of return. This is because, if the market crashes, that 2% is going to start looking like quite a large number. No fund provider wants to be caught out that way, so they will all low-ball their figures to compensate.
With a unit-linked fund, a certain amount of money is put into a pot for you. That money is used to buy shares in various companies, in the hope that their share price will go up. Usually you'll just specify some broad criteria ("I want all my money in Asian shares") and a fund manager will sort out the investment specifics. This approach generally gives a much higher rate of return - e.g. you might get 10% instead of 3%. But it is inherently risky - you might also get -50%.
As pensions actuaries, we want to get the best of both worlds. How can we do this? There are three main approaches (that I'm aware of):
1) Shuffle money from unitised to fixed-interest funds as the member approaches retirement (I mentioned this last time as a good use of SAMs)
2) Invest in an index-linked fund (see footnote*)
3) Invest in a with-profits fund
Technical interlude
A with-profits fund is where you take a fixed interest fund and strap a unitised fund to the roof. The fund provider will keep a pot of money, which it will invest in various companies etc. They will guarantee you a relatively low rate of return (3%, say) on your contribution to this pot.
Then, at the end of every year, they will take a look at how their investments have done. If the investments are doing much better than the guaranteed rate (10%, say), they will tell you "we declare a bonus of (say) 5%". The value of your investment will thus increase by 8%, with the remaining 2% being kept in the pot for a rainy day**.
The point is that, under the with-profits model, the fund provider doesn't need to plan ahead so carefully - if they overdo the bonuses this year, they can just make it up by under-doing it in future years. With-profits funds allow the provider to wing it. They don't have to be as cautious as they are with fixed-interest funds. And that reduced risk translates directly into better rates of return.
What are the characteristics of a with-profit fund? In any given year, there's a limit to how badly they can do (handy if a recession strikes just before you're due to retire). However, in the long term, they won't do as well as unit-linked funds (because they have that extra protection, which must be paid for). They're generally quite a good solution, and fill a very useful niche.
With-profiteering
Now, being smart people, you've probably noticed a gap in the above explanation: how precisely is the bonus calculated? That's an excellent question.
And the answer is very fuzzy. The provider's actuaries will look at the state of the market, look at the state of the fund, and make an educated guess as to what bonus they can afford. There is no hard-and-fast method, and no requirement to document their method. From the point of view of the paying public, the actuaries all go into a dark room and come out a week later saying "right, you're getting 8%". In the words of my generation: What The F***?
Now there are actually good reasons for some of this reticence. An example: many of these calculations use stochastic asset models (see last post) that are just too damn complicated to explain to the general public. Due to the huge pile of regulation hanging anvil-like over our heads, it can often be legally safer to shut up than explain what you're doing. Especially if there's any chance of being creatively misunderstood***.
Regardless of this, the opacity of the system - the secrecy, the lack of information, the argument from authority - definitely promotes suspicion. It's a fricken' invitation to conspiracy theory. I personally would be surprised if the bonus system had ever been abused too badly, just because there are so many people breathing down the average actuary's neck. But you can see how people could come to that conclusion, and there's no way to refute them. Skepticism FAIL.
What can the Institute of Actuaries do about this? Bugger-all, sadly. They don't have the power to compel all these companies to open up, which is really the only way to fix the situation. Hence the rather frantic "oh we're so professional and ethical" song and dance - it's the only thing the poor sods can do to stop their collective street-cred going down the pipe.
Who does have that power? The UK government (in the UK, anyway) and its regulator the Financial Services Authority. Will they act? On existing funds this is very unlikely - it would be too much of a can of worms to open. However, the general trend is towards greater transparency with financial information****. Maybe one day we'll get this mess sorted out.
And the profession will breathe a sigh of relief.
Disclaimer: I am not a subject matter expert, please take all the above with a pinch of salt.
* An index-linked fund guarantees you (say) 1% plus your central bank's base interest rate per year. This removes some of the risk that the buying power of your money will fall (e.g. due to hyperinflation). So although the amount of money you'll get when you retire is uncertain, the number of mars bars you'll be able to buy with it is closer to being fixed.
In terms of returns, index-linked funds are less good than unitised funds, but not nearly as bad as traditional fixed-interest funds. In terms of risks, the order is reversed (as you'd expect).
** There are some complexities (e.g. various bonus types, market value reductions, etc) - this is just an overview not a technical manual.
*** In the UK, the life office Equitable Life went bankrupt at least partly because A) people managed to creatively misunderstand their pensions' guarantee structure, and B) a court decided to hold them to the most expensive interpretation.
**** For example, under the new Solvency II requirements (basically saying how much money an insurer needs to put by in case everything goes wrong at once), the rather complicated models used by each company to calculate their funding requirements must be disclosed. I consider this a very positive step, not least because I want to play around with the models (and possibly blog about them!).
Read the full post
Monday, January 25, 2010
Actuarial theology? Wait a second...
It's funny how different subjects can overlap. This point was forcibly brought home to me by a paper I was reading which compares different "stochastic asset models" (SAMs).
Would you expect a topic like this to tie directly into the philosophy of religion? I didn't. But guess what I found...
Meet SAM
SAMs are bits of mathematics and/or computer code that are used to predict roughly what possible values an asset could hold in future. Actuaries use these to guesstimate how much money you need to put into a pension scheme to get a given payout, and how that money should be invested in the meantime.
As a very simple example, we generally stick money in high-interest (high-risk) assets to start with, and then move it into low-risk (low-interest) assets later on. That's because, as someone approaches retirement, there's less time to recover from disasters - if the market crashes two days before they hit 65, we want to ensure that the piggy-bank is still reasonably full.
Normally the asset ratio is taken to be a simple linear function of time. So for someone retiring at 65, you might say: put min(1,(65-age)/10)*100% of their money into a high-risk fund (e.g. Asian shares) and max(0,(age-55)/10)*100% into a low-risk fund (e.g. government bonds). This is a nice simple linear relationship.
However, it is also a very rough rule of thumb pulled straight out of our actuarial buttocks. The problem is, we don't know how the assets are likely to behave over time, so we can't come up with a more realistic strategy. SAMs are intended to fill that hole.
A SAM will normally be a set of difference equations (or differential equations) with a bit of randomness ("innovations") thrown in for good measure. The equations are intended to codify the relevant bits of economic theory - for example we expect that retail prices will be affected by inflation. The model is then initialised with numbers estimated from the last few years' worth of real-world economic data.
Whoops
The problem* is, these models can sometimes have "interesting" side-effects. For example, the paper I'm reading refers to a type of model which it calls a "random walk variant with alpha-stable innovations". I think I know what that means, but anyway the interesting bit is the following quote:
Yeah, that'd be a problem.
An option is a simple kind of contract - basically a form of insurance. If I run a sawmill, I want to ensure that I can get some logs cheaply no matter what. One approach would be to arrange a "future" contract - a contract which sets a price and time at which I can and must buy the wood.
However, this isn't perfect: what if the price of wood drops massively? In that case I'll be paying good money for something I could buy cheaper elsewhere. Instead of a future, I could arrange an "option" contract, which sets a price and a time at which I can - but don't have to - buy the wood**. This gets me the best of both worlds - no risk of excessive gain or loss. I'll normally have to pay a bit of a premium for the privilege, of course.
These options are vital tools of business, and to be told that they should be infinitely expensive is disturbing. How can that be the case?
Wtf?
As far as I can tell, the logic seems to be: extreme market events are more likely than you'd think. Let's consider a simple case, an option to buy a tonne of wood for £100. We need to work out how much loss the option seller expects to suffer. This will depend on the probability distribution of the lumber's cost C at the option's expiration date.
If we knew for sure that the cost would be £150, we could say P(C=150) = 1. The option seller would then have to buy wood at £150, and sell it to me for only £100. Her loss would be £50, so she'd presumably charge me at least £50 for this (rather useless) option***.
If we guessed that the option price would either go up or down by £30, and that the odds were the same, we could say P(C=50) = P(C=150) = 0.5. If the price went down, I'd be daft to exercise the option, so the seller's expected loss is 0. If the price went up (as it would 50% of the time), the seller would lose $50 as before. £50 × 50% = £25, so the option would probably cost a bit more than £25.
Let's consider a more complicated case:
P(C<=100) = 1/2
P(C=200) = 1/4
P(C=300) = 1/8
P(C=500) = 1/16
P(C=900) = 1/32
...
P(C = 100*(1+2^n)) = 1/4 × 1/2^n for n between 0 and infinity
It's easy to prove that 1/2 + 1/4 + 1/8 + 1/16 + ... = 1, so this is a valid probability distribution.
In this case, what is the option seller's expected loss? It will be:
0×P(C<=100) + (200-100)×P(C=200) + (300-100)×P(C=300) + (500-100)×P(C=500) + ...
This can be rewritten as:
Sum[n = 0 to infinity] { 100 × 2^n × 1/4 × 1/2^n }
= 25 * Sum[n = 0 to inf] { 1 }
But if you add ones together an infinite number of times, you get infinity! This is what's known as a "fat-tail" effect - the "tail" of the distribution is so long and wide that it completely overwhelms the rest of the calculation.
Wager 2.0
From a practical economic perspective, this is a disaster for the model. It's implausible that the value of anything can be infinite - there isn't that much money on the planet.
But I'd like to draw your attention to something. Here we have a model - a belief system of sorts - which, if true, can grant us amazing rewards. All we need to do is act on this belief (e.g. buy an option) and our expected gain goes up to infinity.
Does that remind you of anything?
Pascal's wager is one of the oldest arguments for God. It states that, if you believe in God, you could get to go to Heaven - an infinitely valuable reward - and at worst you'll lose a finite amount of time spent praying etc. If you don't believe, you definitely won't go to Heaven. The "fat tail" of Heaven's infinite goodness is supposed to overwhelm the mere finite rewards you can get from living a happy Godless life.
By the same logic, the infinite return we can expect from buying an option under the alpha-stable model should overwhelm our mere finite uncertainty about whether the model is actually valid. There is no way we can be infinitely sure that the alpha-stable model is wrong - statistics just doesn't work like that. So (in theory) the logic will always hold.
I've given some thought to Pascal's wager, and I think I've found a fair number of holes in it. I haven't thought much about option pricing yet; the same holes may be present there. But in the meantime, a single thought is buzzing round and round in my head:
Should I buy an option?
* OK, so there are all sorts of other problems. But I'll save them for another post.
** Actually this is only one type of option (a European call, to be precise). For more information, see the Wikipedia page.
*** I'm ignoring interest here as it's not really relevant to the point.
Read the full post
Would you expect a topic like this to tie directly into the philosophy of religion? I didn't. But guess what I found...
Meet SAM
SAMs are bits of mathematics and/or computer code that are used to predict roughly what possible values an asset could hold in future. Actuaries use these to guesstimate how much money you need to put into a pension scheme to get a given payout, and how that money should be invested in the meantime.
As a very simple example, we generally stick money in high-interest (high-risk) assets to start with, and then move it into low-risk (low-interest) assets later on. That's because, as someone approaches retirement, there's less time to recover from disasters - if the market crashes two days before they hit 65, we want to ensure that the piggy-bank is still reasonably full.
Normally the asset ratio is taken to be a simple linear function of time. So for someone retiring at 65, you might say: put min(1,(65-age)/10)*100% of their money into a high-risk fund (e.g. Asian shares) and max(0,(age-55)/10)*100% into a low-risk fund (e.g. government bonds). This is a nice simple linear relationship.
However, it is also a very rough rule of thumb pulled straight out of our actuarial buttocks. The problem is, we don't know how the assets are likely to behave over time, so we can't come up with a more realistic strategy. SAMs are intended to fill that hole.
A SAM will normally be a set of difference equations (or differential equations) with a bit of randomness ("innovations") thrown in for good measure. The equations are intended to codify the relevant bits of economic theory - for example we expect that retail prices will be affected by inflation. The model is then initialised with numbers estimated from the last few years' worth of real-world economic data.
Whoops
The problem* is, these models can sometimes have "interesting" side-effects. For example, the paper I'm reading refers to a type of model which it calls a "random walk variant with alpha-stable innovations". I think I know what that means, but anyway the interesting bit is the following quote:
"The alpha-stable distribution [...] is in some ways the most intellectually satisfying possibility. [...However, they are] not satisfactory for the pricing of derivatives because the prices of simple options are theoretically infinite [...]"
Yeah, that'd be a problem.
An option is a simple kind of contract - basically a form of insurance. If I run a sawmill, I want to ensure that I can get some logs cheaply no matter what. One approach would be to arrange a "future" contract - a contract which sets a price and time at which I can and must buy the wood.
However, this isn't perfect: what if the price of wood drops massively? In that case I'll be paying good money for something I could buy cheaper elsewhere. Instead of a future, I could arrange an "option" contract, which sets a price and a time at which I can - but don't have to - buy the wood**. This gets me the best of both worlds - no risk of excessive gain or loss. I'll normally have to pay a bit of a premium for the privilege, of course.
These options are vital tools of business, and to be told that they should be infinitely expensive is disturbing. How can that be the case?
Wtf?
As far as I can tell, the logic seems to be: extreme market events are more likely than you'd think. Let's consider a simple case, an option to buy a tonne of wood for £100. We need to work out how much loss the option seller expects to suffer. This will depend on the probability distribution of the lumber's cost C at the option's expiration date.
If we knew for sure that the cost would be £150, we could say P(C=150) = 1. The option seller would then have to buy wood at £150, and sell it to me for only £100. Her loss would be £50, so she'd presumably charge me at least £50 for this (rather useless) option***.
If we guessed that the option price would either go up or down by £30, and that the odds were the same, we could say P(C=50) = P(C=150) = 0.5. If the price went down, I'd be daft to exercise the option, so the seller's expected loss is 0. If the price went up (as it would 50% of the time), the seller would lose $50 as before. £50 × 50% = £25, so the option would probably cost a bit more than £25.
Let's consider a more complicated case:
P(C<=100) = 1/2
P(C=200) = 1/4
P(C=300) = 1/8
P(C=500) = 1/16
P(C=900) = 1/32
...
P(C = 100*(1+2^n)) = 1/4 × 1/2^n for n between 0 and infinity
It's easy to prove that 1/2 + 1/4 + 1/8 + 1/16 + ... = 1, so this is a valid probability distribution.
In this case, what is the option seller's expected loss? It will be:
0×P(C<=100) + (200-100)×P(C=200) + (300-100)×P(C=300) + (500-100)×P(C=500) + ...
This can be rewritten as:
Sum[n = 0 to infinity] { 100 × 2^n × 1/4 × 1/2^n }
= 25 * Sum[n = 0 to inf] { 1 }
But if you add ones together an infinite number of times, you get infinity! This is what's known as a "fat-tail" effect - the "tail" of the distribution is so long and wide that it completely overwhelms the rest of the calculation.
Wager 2.0
From a practical economic perspective, this is a disaster for the model. It's implausible that the value of anything can be infinite - there isn't that much money on the planet.
But I'd like to draw your attention to something. Here we have a model - a belief system of sorts - which, if true, can grant us amazing rewards. All we need to do is act on this belief (e.g. buy an option) and our expected gain goes up to infinity.
Does that remind you of anything?
Pascal's wager is one of the oldest arguments for God. It states that, if you believe in God, you could get to go to Heaven - an infinitely valuable reward - and at worst you'll lose a finite amount of time spent praying etc. If you don't believe, you definitely won't go to Heaven. The "fat tail" of Heaven's infinite goodness is supposed to overwhelm the mere finite rewards you can get from living a happy Godless life.
By the same logic, the infinite return we can expect from buying an option under the alpha-stable model should overwhelm our mere finite uncertainty about whether the model is actually valid. There is no way we can be infinitely sure that the alpha-stable model is wrong - statistics just doesn't work like that. So (in theory) the logic will always hold.
I've given some thought to Pascal's wager, and I think I've found a fair number of holes in it. I haven't thought much about option pricing yet; the same holes may be present there. But in the meantime, a single thought is buzzing round and round in my head:
Should I buy an option?
* OK, so there are all sorts of other problems. But I'll save them for another post.
** Actually this is only one type of option (a European call, to be precise). For more information, see the Wikipedia page.
*** I'm ignoring interest here as it's not really relevant to the point.
Read the full post
Sunday, January 03, 2010
Saturday, January 02, 2010
Hierarchies of debate
I've just read an interesting post by a guy called Daniel Loxton on James Randi's accidental endorsement of pseudoscience.
I agree with Loxton that only informed debate is useful in scientific matters. But I think his hierarchy of debate as written is a bit too close to an argument from authority. I'd phrase it slightly differently...
1) If someone has clearly done less reading up on the subject than me (e.g. someone saying that evolution cannot add new information, which I know from studying information theory is complete mince), I will vocally disagree with them.
2) If someone has clearly done more reading than me (e.g. they're a Professor of Atmospheric Dynamics at a reputable uni), I will quietly challenge them in the hope that they'll help me improve my understanding. If they appear unable to do so then I'll jump to step #4.
3) If someone appears to have roughly the same level of knowledge as me, I'll aim to trade information with them via debate until we can reach a consensus.
4) If I can't judge whether someone is more or less competent than me, I'll go away and read basic science textbooks until I can judge. (Or I'll shut up. But I don't like shutting up.)
I'm currently at step #4 with my dad on climate change. He's normally more scientifically literate than me, but he's also an AGW skeptic, which seems to go against the consensus. I think it's likely he's just been getting information from dodgy third-hand sources, but I've been getting my information at third hand as well so I can't really protest about that. Yet. I'm working my way through the IPCC report as we speak, and looking up references where possible.
I'm at step #2 with my actuarial science course notes. I have a suspicion that basically the entire financial economics community is taking an approach of "these are the kind of mathematical models we know how to use, therefore we'll assume finance really behaves like that". But I need to do a lot more background reading and questioning, in a state of humility, before I can say for sure.
I consider myself to be at step #1 with the "does God exist" debate, but only because the majority of theists seem to be stuck at step #4 (without the shutting up part). There are some with better knowledge than me, but they tend to be like Henry Neufeld, who freely admits that he believes in God on essentially non-rational grounds.
What hierarchies do you guys use?
Read the full post
I agree with Loxton that only informed debate is useful in scientific matters. But I think his hierarchy of debate as written is a bit too close to an argument from authority. I'd phrase it slightly differently...
1) If someone has clearly done less reading up on the subject than me (e.g. someone saying that evolution cannot add new information, which I know from studying information theory is complete mince), I will vocally disagree with them.
2) If someone has clearly done more reading than me (e.g. they're a Professor of Atmospheric Dynamics at a reputable uni), I will quietly challenge them in the hope that they'll help me improve my understanding. If they appear unable to do so then I'll jump to step #4.
3) If someone appears to have roughly the same level of knowledge as me, I'll aim to trade information with them via debate until we can reach a consensus.
4) If I can't judge whether someone is more or less competent than me, I'll go away and read basic science textbooks until I can judge. (Or I'll shut up. But I don't like shutting up.)
I'm currently at step #4 with my dad on climate change. He's normally more scientifically literate than me, but he's also an AGW skeptic, which seems to go against the consensus. I think it's likely he's just been getting information from dodgy third-hand sources, but I've been getting my information at third hand as well so I can't really protest about that. Yet. I'm working my way through the IPCC report as we speak, and looking up references where possible.
I'm at step #2 with my actuarial science course notes. I have a suspicion that basically the entire financial economics community is taking an approach of "these are the kind of mathematical models we know how to use, therefore we'll assume finance really behaves like that". But I need to do a lot more background reading and questioning, in a state of humility, before I can say for sure.
I consider myself to be at step #1 with the "does God exist" debate, but only because the majority of theists seem to be stuck at step #4 (without the shutting up part). There are some with better knowledge than me, but they tend to be like Henry Neufeld, who freely admits that he believes in God on essentially non-rational grounds.
What hierarchies do you guys use?
Read the full post
Thursday, November 19, 2009
Just my opinion
Anyone who uses the word "quantum" in a serious discussion, and who has never solved Schroedinger's equation, should be stuck in a box with a radioisotope, a neutron-triggered poison dispenser, and an introductory textbook on quantum mechanics. This is what we call "incentive to learn".
First nominee for this treatment: Deepak Chopra. Anyone got their own preferred woo-merchant?
Read the full post
First nominee for this treatment: Deepak Chopra. Anyone got their own preferred woo-merchant?
Read the full post
Tuesday, November 03, 2009
Not Sophistication
Yeah.
So.
This is kinda embarrassing...
How can I put this? Some people are alcoholics. Some people snort cocaine. Some people visit prostitutes. Some people have scary fetishes.
I'm afraid my personal addiction is not nearly as socially acceptable.
Yup, I'm a Dungeons and Dragons webcomic freak.
I swear I never saw it coming. It started off small, just a little bit of Order of the Stick when I was feeling bored. Then a gamer friend landed up in hospital, and I bought him some OOTS books to keep him company, and... I couldn't resist their siren song. I fell into depravity like an expensive, beautifully-painted character model onto a stone floor*.
Next it was Goblins. But hey, two comics ain't so bad. I could give it up any time I liked.
I dabbled with Irregular Webcomic, but quickly moved on to harder drugs. I tasted Looking For Group, and life was sweet for a while.
But then I discovered Darths and Droids and DM Of The Rings. Webcomics about fantasy films, in the style of D&D adventures? I think I've hit bottom here. It's time to admit: I need help.
The really irritating thing is that I don't even play Dungeons and Dragons.
* It's well-known that the force of impact of a dropped model is proportional, not to the height of the table it's knocked off, but to how annoyed you'll be if it gets broken.
Read the full post
So.
This is kinda embarrassing...
How can I put this? Some people are alcoholics. Some people snort cocaine. Some people visit prostitutes. Some people have scary fetishes.
I'm afraid my personal addiction is not nearly as socially acceptable.
Yup, I'm a Dungeons and Dragons webcomic freak.
I swear I never saw it coming. It started off small, just a little bit of Order of the Stick when I was feeling bored. Then a gamer friend landed up in hospital, and I bought him some OOTS books to keep him company, and... I couldn't resist their siren song. I fell into depravity like an expensive, beautifully-painted character model onto a stone floor*.
Next it was Goblins. But hey, two comics ain't so bad. I could give it up any time I liked.
I dabbled with Irregular Webcomic, but quickly moved on to harder drugs. I tasted Looking For Group, and life was sweet for a while.
But then I discovered Darths and Droids and DM Of The Rings. Webcomics about fantasy films, in the style of D&D adventures? I think I've hit bottom here. It's time to admit: I need help.
The really irritating thing is that I don't even play Dungeons and Dragons.
* It's well-known that the force of impact of a dropped model is proportional, not to the height of the table it's knocked off, but to how annoyed you'll be if it gets broken.
Read the full post
Sunday, October 18, 2009
Sophistication
You know that guy? The one who is always in great shape despite apparently living off fast food. The one who always gets top marks despite apparently never revising for exams. The one who can drive, skate, ski, swim, fight, play every card game known to man, and all without ever seeming to break a sweat.
Everyone knows someone like this (guy or girl). After years of wondering how the blazes they do all that and still look so laid back, I've come to a conclusion.
They're frauds.
Sure, they may have a slightly broader range of talents than the average bloke. Sure, they possibly started out with slightly better strength and dexterity than us mere mortals. But there is no level of innate ability that could set them that high above the rest of us.
I think that, for every burger you see being eaten, there's an hour in the gym that you never find out about. For every cakewalk of an exam, there are many frantic hours of secret preparation. For every activity that they're "just naturally good" at, they have undoubtedly spent time preparing and training.
This is reassuring. It means that anyone can do what that guy does. Anyone can upgrade themselves to the status of god among men. It just takes a lot of work. Specifically, it takes a lot of work without any sort of immediate reward.
I now have a goal in life. Firstly, to practice the discipline required for this sort of long-term training plan. Secondly, to develop the sophistication required to STFU.
I will make better progress towards both of these goals if I get a good eight hours sleep. Goodnight.
Read the full post
Everyone knows someone like this (guy or girl). After years of wondering how the blazes they do all that and still look so laid back, I've come to a conclusion.
They're frauds.
Sure, they may have a slightly broader range of talents than the average bloke. Sure, they possibly started out with slightly better strength and dexterity than us mere mortals. But there is no level of innate ability that could set them that high above the rest of us.
I think that, for every burger you see being eaten, there's an hour in the gym that you never find out about. For every cakewalk of an exam, there are many frantic hours of secret preparation. For every activity that they're "just naturally good" at, they have undoubtedly spent time preparing and training.
This is reassuring. It means that anyone can do what that guy does. Anyone can upgrade themselves to the status of god among men. It just takes a lot of work. Specifically, it takes a lot of work without any sort of immediate reward.
I now have a goal in life. Firstly, to practice the discipline required for this sort of long-term training plan. Secondly, to develop the sophistication required to STFU.
I will make better progress towards both of these goals if I get a good eight hours sleep. Goodnight.
Read the full post
Bwahahaha
I've been trying to cut back on my public use of mad-scientist laughter. Despite its proven stress-relieving effect, it does disturb my co-workers somewhat.
To compensate, I intend to use my blog as a gloating platform. I consider this to be ethically acceptable on three grounds:
1) My blog, my rules (incidentally, the new dress code round here is "winged monkey")
2) Blog-reading is strictly voluntary - you fools chose to read this garbage
3) It's not like you can do anything about it anyway
Those preliminaries out the way, I would just like to say I'm very happy. And boastful. But mostly happy.
About four months back now, I got bounced up to a different part of England - I'm sure I've whined about it previously. My temporary home is a hotel in a little sea town that wishes it was Las Vegas. For generations, anyone with any brains or talent has been escaping from this dump*, and the result is reminiscent of Innsmouth without the successful fishing industry.
Coming to terms with the mindless tedium presents an interesting challenge. As a partial solution (because there's only so many books you can cram in a suitcase - believe me I know) I've taken up Taekwondo. I've previously done Karate, but they don't have a club for that within walking distance of the hotel, so what the hey.
I've actually been really enjoying it. This is the first martial arts training I've done in about three years, and it's been good to feel the old skills starting to come back (plus a few new ones). It's been going so well that I've been rather looking forward to the grading, which was earlier today.
Ah, the grading. I used to think the sweetest words you could hear in a grading were "you've passed". It turns out I was wrong. The sweetest words are "we've decided to let this student skip a belt". I've jumped yellow-belt entirely and gone straight to yellow-with-green-tag. Who thinks up these colours?
Normally I would take this with a large dose of humility - until this morning I was the only adult white-belt at the club, so they could have just been letting me catch up with my "peer group". However, when they were handing out the new belts, it transpired that they did not have a green-tag belt with them, which suggests that this was a spur-of-the-moment decision based on my performance in the grading.
Quite apart from providing me with excellent bragging rights, this episode also highlights the unreasonable effectiveness of the human brain (yes, even mine). Once connections are made, they tend to stay made. Once a skill is learned, it persists far beyond its anticipated sell-by date.
The moral: never be afraid to spend a bit of time learning a new skill, or polishing an old one. I'm going to be leaving this area in December, so I'll only have had a few months with this club. But even this short few months has been good for me. And I know that, next time I decide to take up a martial art, it'll be easier than ever.
* This is unduly harsh. There are many good, intelligent people here - I work with a bunch of them. But the social agenda seems to be completely controlled by people who think fart jokes are the height of humour, big flashing lights make a place look modern and tasteful, psychics can solve all your problems, and Sophistication is an island in Greece.
Read the full post
To compensate, I intend to use my blog as a gloating platform. I consider this to be ethically acceptable on three grounds:
1) My blog, my rules (incidentally, the new dress code round here is "winged monkey")
2) Blog-reading is strictly voluntary - you fools chose to read this garbage
3) It's not like you can do anything about it anyway
Those preliminaries out the way, I would just like to say I'm very happy. And boastful. But mostly happy.
About four months back now, I got bounced up to a different part of England - I'm sure I've whined about it previously. My temporary home is a hotel in a little sea town that wishes it was Las Vegas. For generations, anyone with any brains or talent has been escaping from this dump*, and the result is reminiscent of Innsmouth without the successful fishing industry.
Coming to terms with the mindless tedium presents an interesting challenge. As a partial solution (because there's only so many books you can cram in a suitcase - believe me I know) I've taken up Taekwondo. I've previously done Karate, but they don't have a club for that within walking distance of the hotel, so what the hey.
I've actually been really enjoying it. This is the first martial arts training I've done in about three years, and it's been good to feel the old skills starting to come back (plus a few new ones). It's been going so well that I've been rather looking forward to the grading, which was earlier today.
Ah, the grading. I used to think the sweetest words you could hear in a grading were "you've passed". It turns out I was wrong. The sweetest words are "we've decided to let this student skip a belt". I've jumped yellow-belt entirely and gone straight to yellow-with-green-tag. Who thinks up these colours?
Normally I would take this with a large dose of humility - until this morning I was the only adult white-belt at the club, so they could have just been letting me catch up with my "peer group". However, when they were handing out the new belts, it transpired that they did not have a green-tag belt with them, which suggests that this was a spur-of-the-moment decision based on my performance in the grading.
Quite apart from providing me with excellent bragging rights, this episode also highlights the unreasonable effectiveness of the human brain (yes, even mine). Once connections are made, they tend to stay made. Once a skill is learned, it persists far beyond its anticipated sell-by date.
The moral: never be afraid to spend a bit of time learning a new skill, or polishing an old one. I'm going to be leaving this area in December, so I'll only have had a few months with this club. But even this short few months has been good for me. And I know that, next time I decide to take up a martial art, it'll be easier than ever.
* This is unduly harsh. There are many good, intelligent people here - I work with a bunch of them. But the social agenda seems to be completely controlled by people who think fart jokes are the height of humour, big flashing lights make a place look modern and tasteful, psychics can solve all your problems, and Sophistication is an island in Greece.
Read the full post
Sunday, September 27, 2009
Where did they go?
I'm not a good programmer. I'm naturally quite techie, but I've never really had the patience to sit down and make great art with my PC. The skills I have are those I was able to absorb from those in my immediate vicinity - replication not initiation.
That said, I do enjoy absorbing the culture of tech - the story of mel, the jargon file, esoteric languages, and the old tales of the MIT AI lab and Xerox Park. I read books by Neal Stephenson and William Gibson. Buried somewhere on my hard drive is the complete archives of Phrack mag, although I barely understand half of it (typically the obsolete half - yay analog phreaking!).
And one of the things I pick up from these shards of geekiness is a sense of wistfulness. They talk about the September that never ended. They talk about the AI Winter. They refer to newsgroups and communities of unsurpassed elegance and sophistication, that now no longer exist. There's a sense of stumbling across the forgotten artifacts of some lost higher civilisation. Once heroes and wizards strode the Earth; now there are only echoes.
Where did they go? What happened to the cypherpunk generation? Did they all quit programming, or wind up in wage-slave jobs that crushed their creativity, or die of drug overdoses, or get locked up for not respecting someone's lack of respect for computer security? Have the elves passed into the West, making way for the Age of Man?
I hope not. I like to think that, somewhere, these ideals hold on. Somewhere, in a hidden mailing list, on a firewalled server, carried by a stream of encrypted emails spliced into innocuous data, the crypto-anarchist dream lives on. It's just waiting to be found, locked behind doors that cry out for the right key. So what if the key in question is 8192-bit?
Maybe I'm deluding myself. Maybe the cypherpunk movement just died out, faded back into oblivion. It would be a poorer world if that were so, but the world has no responsibility to respect our desires.
But I allow myself this one dream. And in consolation for the lack of evidence, I hold this thought tightly:
If they couldn't hide themselves from people like me, they wouldn't be worth admiring...
Read the full post
That said, I do enjoy absorbing the culture of tech - the story of mel, the jargon file, esoteric languages, and the old tales of the MIT AI lab and Xerox Park. I read books by Neal Stephenson and William Gibson. Buried somewhere on my hard drive is the complete archives of Phrack mag, although I barely understand half of it (typically the obsolete half - yay analog phreaking!).
And one of the things I pick up from these shards of geekiness is a sense of wistfulness. They talk about the September that never ended. They talk about the AI Winter. They refer to newsgroups and communities of unsurpassed elegance and sophistication, that now no longer exist. There's a sense of stumbling across the forgotten artifacts of some lost higher civilisation. Once heroes and wizards strode the Earth; now there are only echoes.
Where did they go? What happened to the cypherpunk generation? Did they all quit programming, or wind up in wage-slave jobs that crushed their creativity, or die of drug overdoses, or get locked up for not respecting someone's lack of respect for computer security? Have the elves passed into the West, making way for the Age of Man?
I hope not. I like to think that, somewhere, these ideals hold on. Somewhere, in a hidden mailing list, on a firewalled server, carried by a stream of encrypted emails spliced into innocuous data, the crypto-anarchist dream lives on. It's just waiting to be found, locked behind doors that cry out for the right key. So what if the key in question is 8192-bit?
Maybe I'm deluding myself. Maybe the cypherpunk movement just died out, faded back into oblivion. It would be a poorer world if that were so, but the world has no responsibility to respect our desires.
But I allow myself this one dream. And in consolation for the lack of evidence, I hold this thought tightly:
If they couldn't hide themselves from people like me, they wouldn't be worth admiring...
Read the full post
Sunday, September 20, 2009
How To Learn French
1) Work your way through Wikibooks: French
2) Download a freeware French-English dictionary (I'm usingFreedict - open-source but not very user-friendly).
3) Download a good book here and ici, and read them side by side.
4) Tune in to web radio (alternatively pick a DVD, and set the language to French and the subtitles to English).
It's amazing how long you can spend on the internet and still not use it to its full potential.
Read the full post
2) Download a freeware French-English dictionary (I'm usingFreedict - open-source but not very user-friendly).
3) Download a good book here and ici, and read them side by side.
4) Tune in to web radio (alternatively pick a DVD, and set the language to French and the subtitles to English).
It's amazing how long you can spend on the internet and still not use it to its full potential.
Read the full post
Monday, August 17, 2009
An example thereof
In my last post I explained why it was a good thing to be skeptical before passing ideas on.
Now, I've just come across a great reference to "the world's oldest apocalypse prediction":
This reference is sourced to Isaac Asimov's Book of Facts. The problem is, I can't substantiate it. I don't have a copy of the Book of Facts handy and, even if I did, I would need to know more about the clay tablet in question before I could trust Mr Asimov's word on this.
So if I were to use this quote in any sort of serious discussion, I would need to accompany it with a shot of skeptical "penicillin". I would have to make my friends aware that I could not stake my life on the it being accurate. This would be boring and long-winded.
The only alternative is to try to track down the tablet in question online. This is not proving easy: googling for the translated text just finds thousands of people who have quite clearly copied it straight out of the Book of Facts. This is not corroboration.
So I need to dig deeper. With a bit of effort, I'll be able to figure out how the Assyrian research community organises its information, which should give me some idea of where to find this particular tablet. So far I've come across the Cuneiform Digital Library Initiative (not so helpful as it doesn't give translations) and the Neo-Assyrian Text Corpus Project (which appears to be defunct).
Beyond that, I may have to gatecrash the local university library. Watch this space.
It may take a while to learn the truth here. Heck, I might actually need to learn Assyrian to track down the tablet (or to demonstrate that it probably doesn't exist). I am unlikely to go that far. But the time I do spend on this exercise will be time well used - a tithe spent on improving the information available to the community as a whole.
Read the full post
Now, I've just come across a great reference to "the world's oldest apocalypse prediction":
"Our earth is degenerate in these latter days. There are signs that the world is speedily coming to an end. Bribery and corruption are common."
- Assyrian clay tablet, circa 2800BC
This reference is sourced to Isaac Asimov's Book of Facts. The problem is, I can't substantiate it. I don't have a copy of the Book of Facts handy and, even if I did, I would need to know more about the clay tablet in question before I could trust Mr Asimov's word on this.
So if I were to use this quote in any sort of serious discussion, I would need to accompany it with a shot of skeptical "penicillin". I would have to make my friends aware that I could not stake my life on the it being accurate. This would be boring and long-winded.
The only alternative is to try to track down the tablet in question online. This is not proving easy: googling for the translated text just finds thousands of people who have quite clearly copied it straight out of the Book of Facts. This is not corroboration.
So I need to dig deeper. With a bit of effort, I'll be able to figure out how the Assyrian research community organises its information, which should give me some idea of where to find this particular tablet. So far I've come across the Cuneiform Digital Library Initiative (not so helpful as it doesn't give translations) and the Neo-Assyrian Text Corpus Project (which appears to be defunct).
Beyond that, I may have to gatecrash the local university library. Watch this space.
It may take a while to learn the truth here. Heck, I might actually need to learn Assyrian to track down the tablet (or to demonstrate that it probably doesn't exist). I am unlikely to go that far. But the time I do spend on this exercise will be time well used - a tithe spent on improving the information available to the community as a whole.
Read the full post
Subscribe to:
Posts (Atom)
