30 December 2008
Can Google Survive Its Success?
As I understand (admittedly dimly) Google's success is based on an algorithm that does a particularly good job ranking search results based upon how many other web pages link to it. But what happens now that no one finds anything on the web other than through Google? Now Google's algorithm must basically be to rank pages in the order that Google ranks them. I understand that now they're concentrating on getting every web page in the world into their servers multiple times, but that seems like a pretty low-return game.
26 December 2008
Happy Boxing Day
Today is Boxing Day, the day when, um, er, I have no idea (Ma'nish ta'na...).
Is it:
a) The day baby Jesus kicked the Wise Men's asses?
b) The day in England on which the servants are expected to wrap their presents from their masters?
c) A day for organising one's closet? Or,
d) A day in which the English go to "Pantomimes," which they watch from theatre boxes.
Whatever it is, let's all enjoy it. Holidays need no excuses.
Is it:
a) The day baby Jesus kicked the Wise Men's asses?
b) The day in England on which the servants are expected to wrap their presents from their masters?
c) A day for organising one's closet? Or,
d) A day in which the English go to "Pantomimes," which they watch from theatre boxes.
Whatever it is, let's all enjoy it. Holidays need no excuses.
25 December 2008
Christmas Conundrum
Trying to run a hierarchical linear model, I've discovered that my dependent variable, Return on Assets, is leptokurtotic. I can improve it somewhat by running a full model on the raw data, analyzing the residuals and dropping some outliers, but ROA is still leptokurtotic. A logit transform, which is supposed to reduce kurtosis, just makes things worse. It turns out that ROA, which is a very common dependent variable in Strategic Management, is almost always kurtotic because it violates the assumption of proportionality; that is, the relationship between net income and total assets is non-linear.
What is kurtosis? Here's a picture of leptokurtotic data plotted against a normal curve:
What makes it kurtotic is that it is "peakier" than the normal curve. Why should we care? Because the most common statistical methods assume that the residuals of the dependent variable (how far each point is from the predicted value) are normally distributed. If they aren't, then the estimate of how likely it is that a particular result found in a particular sample is chance rather than real is unreliable. Since that's all that statistics do, that's a problem. So, for example, if researchers assume that headaches are distributed normally in the population but they're not, then a treatment that seems effective might not be.
More to the point, some economists think that Long Term Capital Management failed because its models assumed that certain financial measures, like Return on Assets, were normally distributed when they're not. At least one economist has argued that the Black-Scholes option pricing model, which is related to the LTCM model is also wrong for the same reason. Since Black-Scholes and other, related, models are the basis for most modern finance, that implies that people pricing the risk of various financial instruments without much of a history might have assumed that Return on Assets, etc., are normally distributed. Because, contrary to this assumption, ROA is leptokurtotic, the models might have overestimated the expected return on investment, underestimated the risk and mispriced the instruments.
In other words, the fact that ROA is leptokurtotic is a possible explanation, and a more satisfying explanation than most out there, for the sub-prime mortgage and CDS implosion.
What is kurtosis? Here's a picture of leptokurtotic data plotted against a normal curve:
What makes it kurtotic is that it is "peakier" than the normal curve. Why should we care? Because the most common statistical methods assume that the residuals of the dependent variable (how far each point is from the predicted value) are normally distributed. If they aren't, then the estimate of how likely it is that a particular result found in a particular sample is chance rather than real is unreliable. Since that's all that statistics do, that's a problem. So, for example, if researchers assume that headaches are distributed normally in the population but they're not, then a treatment that seems effective might not be.
More to the point, some economists think that Long Term Capital Management failed because its models assumed that certain financial measures, like Return on Assets, were normally distributed when they're not. At least one economist has argued that the Black-Scholes option pricing model, which is related to the LTCM model is also wrong for the same reason. Since Black-Scholes and other, related, models are the basis for most modern finance, that implies that people pricing the risk of various financial instruments without much of a history might have assumed that Return on Assets, etc., are normally distributed. Because, contrary to this assumption, ROA is leptokurtotic, the models might have overestimated the expected return on investment, underestimated the risk and mispriced the instruments.
In other words, the fact that ROA is leptokurtotic is a possible explanation, and a more satisfying explanation than most out there, for the sub-prime mortgage and CDS implosion.
23 December 2008
Recommendations
Of all the different ways of listening to music on the internet, I've listening to Pandora. Basically, you choose a genre, artist or song you like and Pandora picks out features of that song that it looks for in other songs. At the moment, I've just started a "Fred Smith" channel and Pandora is looking for songs that have "folk rock qualities, extensive vamping, acoustic sonority, major key tonality and a breathy male lead vocalist." Well worth suffering through their affectation that they've identified musical DNA.
A Pirate Premium?
Does anyone know why, over the last week or so, the WTI Cushing price for a barrel of oil has started trading at a $5-$10 dollar discount to the other spot markets?
Festivus Comes Early
Via Ann Althouse, I see that the New York Times printed a letter to the editor from the Mayor of Paris lamenting the anti-democratic nature of Caroline Kennedy's possible appointment to the Senate. It reeks, he implies, of American decline.
The letter is a fake.
Now, on to the feats of strength.
The letter is a fake.
Now, on to the feats of strength.
21 December 2008
Timing The Surge
When it does become time to judge George W. Bush's presidency, one big question will concern the timing of the surge. If it could have been done successfully earlier, than it should have been. If it had to wait for AQiM to prove to the Sunnis what vicious idiots they were, then it couldn't have been done much sooner.
19 December 2008
Rick Warren At The Inauguration
First thought: If McCain had won and invited Warren, the progressive left would have told us about how this just proves that Obama would have been so much better.
Second thought: For the left, this is just like how it would have been for the right if McCain had invited Rev. Wright to give the invocation.
Third thought: That's exactly what McCain would have done.
Second thought: For the left, this is just like how it would have been for the right if McCain had invited Rev. Wright to give the invocation.
Third thought: That's exactly what McCain would have done.
What Does Democracy Do?
There are three possible justifications for democracy (including representative democracy):
There's some truth to all of this, but each thesis has different policy implications. If the kindergarten thesis is right -- given the chance, the American people will do the right thing -- then we care about information flow and informed voters. We want to think about things like poll taxes or literacy tests, and we want to force the candidates to make public their specific plans. We care about campaign promises the same way we care about product warranties and fitness for a particular purpose. Parliamentary systems tend to follow this model.
In the Jeffersonian thesis, we're more interested in the people running for office than in what they plan to do. We want exposure to the candidates personality so that we can judge what kind of people they are, rather than their specific plans for specific problems. This is a pretty good positive description of the current system in which we make the candidates spend two years working long hard hours debating, eating corn dogs, giving speeches, going on Letterman, writing (or putting their name to) articles in Foreign Affairs, etc.
In the black box model, all we need as a decision. We don't have to know or care how that decision was arrived at. Random works fine, so long as all the actors agree to be bound by the decision. the black box thesis doesn't have much to say about the campaign, but it loves the Electoral College in which even a random and ambiguous close vote is translated into a clear win.
I find, as I get older, that I become more attracted to the black box thesis. Given that we need to have a president, for example, and that he needs to be the clear winner, an election is as good a way to choose him as any and better than most. At that point, retrospective sense making and our psychological need to see ourselves in control of our fate will kick in and we'll all agree that he is the legitimate leader, chosen by the Jeffersonian electorate. (And, if push comes to shove, G-d won't let anything too bad happen to us.)
(We can also add the Bismark corollary, which says that it doesn't really matter how we choose our leaders because G-d's going to take care of us anyway, but I assume that it's obvious why the less we talk about that, the better.)
- The people, corporately, make good decisions (which I'll call the "kindergarten thesis");
- Election to office lends legitimacy from the sovereign people to our temporary leaders (the "Jeffersonian theis"); or
- We need to get our leaders somewhere and voting beats accident of birth or force of arms (the "black box thesis").
There's some truth to all of this, but each thesis has different policy implications. If the kindergarten thesis is right -- given the chance, the American people will do the right thing -- then we care about information flow and informed voters. We want to think about things like poll taxes or literacy tests, and we want to force the candidates to make public their specific plans. We care about campaign promises the same way we care about product warranties and fitness for a particular purpose. Parliamentary systems tend to follow this model.
In the Jeffersonian thesis, we're more interested in the people running for office than in what they plan to do. We want exposure to the candidates personality so that we can judge what kind of people they are, rather than their specific plans for specific problems. This is a pretty good positive description of the current system in which we make the candidates spend two years working long hard hours debating, eating corn dogs, giving speeches, going on Letterman, writing (or putting their name to) articles in Foreign Affairs, etc.
In the black box model, all we need as a decision. We don't have to know or care how that decision was arrived at. Random works fine, so long as all the actors agree to be bound by the decision. the black box thesis doesn't have much to say about the campaign, but it loves the Electoral College in which even a random and ambiguous close vote is translated into a clear win.
I find, as I get older, that I become more attracted to the black box thesis. Given that we need to have a president, for example, and that he needs to be the clear winner, an election is as good a way to choose him as any and better than most. At that point, retrospective sense making and our psychological need to see ourselves in control of our fate will kick in and we'll all agree that he is the legitimate leader, chosen by the Jeffersonian electorate. (And, if push comes to shove, G-d won't let anything too bad happen to us.)
18 December 2008
And You Think We Have Problems
This has got to be the weirdest comment thread I've ever read, including on Screw Loose Change.
No Logic Please, We're American
OJ points us to an article arguing that the Europeans hate us because of our freedoms. Of course, the terrorists hate us because of our freedoms, too. So, as we all suspected, Europeans are terrorists.
14 December 2008
"So what if a guy threw his shoe at me,"
George W. Bush is, in the end, probably not a great president. He is a little too dismissive of the bully pulpit to qualify as great, a characteristic failing of his family. Supposedly, his grandmother once complained that George H.W. Bush was talking about himself too much in one of his campaigns and this reticence, odd in a family that gave us two presidents, seems to have affected W. as well. It is also, I suspect, partly how Laura likes it and partly an entirely laudable effort not to be Bill Clinton.
He has been, however, the perfect president for me. For better or worse, if I were designing a president, I'd end up with George W. Bush. And I for one will dearly miss a president who's reaction to having a shoe thrown at him is "So what?"
He has been, however, the perfect president for me. For better or worse, if I were designing a president, I'd end up with George W. Bush. And I for one will dearly miss a president who's reaction to having a shoe thrown at him is "So what?"
12 December 2008
Today's "Don't We Know Too Much About Each Other For Strangers" Quiz
James Lileks points us to a mystery cartoon that uses the phrase "A drug on the market."
Name the only other context in which I've heard this phrase.
Well, I guess I don't know you people as well as I thought:
Name the only other context in which I've heard this phrase.
Well, I guess I don't know you people as well as I thought:
At another table a very well-dressed WOMAN talks to a MOOR. She has a bracelet on her wrist. No other jewelry.Casablanca, 1942.WOMAN
But can't you make it just a little more? Please.
MOOR
I'm sorry, Madam, but diamonds are a drug on the market. Everybody sells diamonds. There are diamonds everywhere. Two thousand, four hundred.
WOMAN
All right.
11 December 2008
What's Really Disturbing About This:
This is becoming a terrible week for the US newspaper industry. On Monday, the Tribune Company, which owns the Chicago Tribune and the Los Angeles Times, filed for bankruptcy. The New York Times Company followed by saying it might mortgage its Renzo Piano-designed headquarters building by Times Square to reduce debt.is that it's from the Financial Times. It is, of course, from an article about the importance of the rock-solid reporting we get from The New York Times, the Wall Street Journal and the Financial Times.
There Are Only Three Successful Arguments In American Politics
1. God wants us to do this.
2. It's for the children.
3. This might lead to war.
(I would link to the Daily Duck piece that got me going down this track, but I can't reach the Daily Duck at the moment.)
2. It's for the children.
3. This might lead to war.
(I would link to the Daily Duck piece that got me going down this track, but I can't reach the Daily Duck at the moment.)
09 December 2008
A Conundrum
One months ago, the yield on the 5-year Treasury note was 2.51%. Today the yield is 1.61%. Over the same period, the price of a Credit Default Swap for the 5-year Treasury, basically insurance against the Treasury not paying off, has risen to $60 per $10,000 insured from $35.70, an increase of 68%. (A year ago, the cost was $8.00 per $10,000.) In other words, the risk has gone up while the yield has gone down. Generally, so much money is flooding into Treasuries that short-term notes are trading at essentially 0, and some have a negative real yield (that is, the interest paid is less than inflation). Anyone want to offer an explanation?
06 December 2008
Peter Responds
With the, ur, "short" version. Hey, I report, you decide:
1) With 38% of the popular vote, Harper wins a strong minority government in October in a field of five parties, four of a leftist bent. He is twelve seats short of a majority. He wins his first confidence motion, but then clumsily issues a budget statement with controversial budget cuts including one that will reduce the funding base of the other parties. Harper has a lot of strengths, but overt petulant partisanship is his Achilles Heel.
2) Completely out of the blue, the three opposition parties (the Greens won no seats), including the separatist-in-principle-but-we-promise-not-until-some-vague-future-date-we-need-not-specify-now Bloc Quebecois that has never been included in government in any way, announce a Liberal/NDP (socialist)coalition with promised Bloc support that moves non-confidence and announces it is ready to govern when the Cons are defeated. They then argue that the Governor-General, the royal representative, is constitutionally obliged to invite them to form a government and refuse Harper's request for an election because we just had one. They are all broke from the last election and only the Tories are in good financial health. Several constitutional experts, who finally get their fifteen minutes of fame after spending many lonely years mastering arcane legal/historical stuff nobody gives a crap about, agree. Pandemonium, because there are only days to the motion. Harper, who controls the parliamentary agenda, adjourns it for a week to Dec 8th.
3) The week is crazy and increasingly very disturbing because the rhetoric moves to overheat and everybody is nakedly partisan at a time the country wants steady-as-she-goes leadership. It becomes apparent the coalition was in the works long before the budget statement. Harper offers to backtrack on the most controversial items, but there is no stopping them now. Trouble is, the Liberal leader is an unpopular wimp who blew the election and resigned a few days later, effective a leadership convention next May that already has several declared candidates out campaigning, so who will the new PM be? It has to be a Liberal because they have by far the most seats, but they have no credible leader. Internal uncertainty and tensions about that surface until it is decided he will anyway, which does the coalition no good at all in popular support;
4) The big issue seems to be whether the GG, who rushes home from a diplomatic sojourn in Europe, will grant the election it is presumed Harper will ask for because it looks like he might cream the leaderless Libs. Never in my life have so many Canadians learned so much about the fabled royal perogative. Brit will know all about this, but basically the Crown does what the Government requests and advises 99.99999% of the time. However, there are theoretical reserve powers to act independantly if Parliament is being subverted or in times of crisis;
5) Then it starts to become clear Harper will not wait for the vote, but will instead ask the GG to prorogue Parliament (formally ending the session which requires royal assent, as opposed to just adjourning for a few weeks which the House decides itself by majority vote) before so everybody can take a Valium until late January. Same wild debate as to whether she is compelled to agree just weeks after the last election and with a theoretical alternative government to call on. Constitutional experts get a second fifteen minutes of fame. The coalition screams about subverting democracy and governing as a dictator, yada, yada. It first looks like Harper's move is a cowardly dealing from the bottom of the deck, but as everybody starts to get alarmed by the turmoil and alternatives, support for it grows because the alternatives (coalition government or another election) are both so unpopular. Everybody, including me, starts thinking: "This is neither fun nor funny anymore."
6) It would take too long to explain, but the mess is terribly divisive and divides the country badly, both politically and regionally. Pundits work 24/7. Disgust for politicians and the process spikes. The public wants cooperation and statesmanship has other things on their minds like keeping their jobs. The coalition really hasn't got its act together because they hate and distrust one another personally as much as they hate the Tories publically.
7) Thursday the GG grants the request to prorogue and the air rushes out of the balloon. National polls start to show strong anti-coalition feeling and Harper's support soars except in Quebec and show he would sweep to a majority. Uh-oh, nice play, Shakespeare. Dissension in the ranks of the left emerge and many say the coalition won't last. If just 12 of the opposition M.P.'s refuse to vote or vote with the Tories, it's game, set, match to Harper, and a few start questionning it publically. Leftist blogs start infighting about the disaster, which of course is all blamed on Tory lies and spin. The coalition starts to descend into buffoonery, such as failing to get the tape of it's leader's address to the nation to TV feed stations in Ottawa in time for specially scheduled national statements. A delicious, much-quoted quip from a radio talk show is that is looks like "Arrogance vs. The Three Stooges".
So, the country is going to take a break for a nice Christmas dinner and nobody has a clue about January. Harper is the clear winner but only in the relative context of an overall bilious disgust with politicians.
Finger-pointing all around, but we of the post-Judd Alliance all know that the real reason for this incredibly risky failed gambit is that conservative governments are offences against progress and the Enlightenment, if not history itself.
David: Actually, I'll put it down to what a bad idea Parliamentary government is.
1) With 38% of the popular vote, Harper wins a strong minority government in October in a field of five parties, four of a leftist bent. He is twelve seats short of a majority. He wins his first confidence motion, but then clumsily issues a budget statement with controversial budget cuts including one that will reduce the funding base of the other parties. Harper has a lot of strengths, but overt petulant partisanship is his Achilles Heel.
2) Completely out of the blue, the three opposition parties (the Greens won no seats), including the separatist-in-principle-but-we-promise-not-until-some-vague-future-date-we-need-not-specify-now Bloc Quebecois that has never been included in government in any way, announce a Liberal/NDP (socialist)coalition with promised Bloc support that moves non-confidence and announces it is ready to govern when the Cons are defeated. They then argue that the Governor-General, the royal representative, is constitutionally obliged to invite them to form a government and refuse Harper's request for an election because we just had one. They are all broke from the last election and only the Tories are in good financial health. Several constitutional experts, who finally get their fifteen minutes of fame after spending many lonely years mastering arcane legal/historical stuff nobody gives a crap about, agree. Pandemonium, because there are only days to the motion. Harper, who controls the parliamentary agenda, adjourns it for a week to Dec 8th.
3) The week is crazy and increasingly very disturbing because the rhetoric moves to overheat and everybody is nakedly partisan at a time the country wants steady-as-she-goes leadership. It becomes apparent the coalition was in the works long before the budget statement. Harper offers to backtrack on the most controversial items, but there is no stopping them now. Trouble is, the Liberal leader is an unpopular wimp who blew the election and resigned a few days later, effective a leadership convention next May that already has several declared candidates out campaigning, so who will the new PM be? It has to be a Liberal because they have by far the most seats, but they have no credible leader. Internal uncertainty and tensions about that surface until it is decided he will anyway, which does the coalition no good at all in popular support;
4) The big issue seems to be whether the GG, who rushes home from a diplomatic sojourn in Europe, will grant the election it is presumed Harper will ask for because it looks like he might cream the leaderless Libs. Never in my life have so many Canadians learned so much about the fabled royal perogative. Brit will know all about this, but basically the Crown does what the Government requests and advises 99.99999% of the time. However, there are theoretical reserve powers to act independantly if Parliament is being subverted or in times of crisis;
5) Then it starts to become clear Harper will not wait for the vote, but will instead ask the GG to prorogue Parliament (formally ending the session which requires royal assent, as opposed to just adjourning for a few weeks which the House decides itself by majority vote) before so everybody can take a Valium until late January. Same wild debate as to whether she is compelled to agree just weeks after the last election and with a theoretical alternative government to call on. Constitutional experts get a second fifteen minutes of fame. The coalition screams about subverting democracy and governing as a dictator, yada, yada. It first looks like Harper's move is a cowardly dealing from the bottom of the deck, but as everybody starts to get alarmed by the turmoil and alternatives, support for it grows because the alternatives (coalition government or another election) are both so unpopular. Everybody, including me, starts thinking: "This is neither fun nor funny anymore."
6) It would take too long to explain, but the mess is terribly divisive and divides the country badly, both politically and regionally. Pundits work 24/7. Disgust for politicians and the process spikes. The public wants cooperation and statesmanship has other things on their minds like keeping their jobs. The coalition really hasn't got its act together because they hate and distrust one another personally as much as they hate the Tories publically.
7) Thursday the GG grants the request to prorogue and the air rushes out of the balloon. National polls start to show strong anti-coalition feeling and Harper's support soars except in Quebec and show he would sweep to a majority. Uh-oh, nice play, Shakespeare. Dissension in the ranks of the left emerge and many say the coalition won't last. If just 12 of the opposition M.P.'s refuse to vote or vote with the Tories, it's game, set, match to Harper, and a few start questionning it publically. Leftist blogs start infighting about the disaster, which of course is all blamed on Tory lies and spin. The coalition starts to descend into buffoonery, such as failing to get the tape of it's leader's address to the nation to TV feed stations in Ottawa in time for specially scheduled national statements. A delicious, much-quoted quip from a radio talk show is that is looks like "Arrogance vs. The Three Stooges".
So, the country is going to take a break for a nice Christmas dinner and nobody has a clue about January. Harper is the clear winner but only in the relative context of an overall bilious disgust with politicians.
Finger-pointing all around, but we of the post-Judd Alliance all know that the real reason for this incredibly risky failed gambit is that conservative governments are offences against progress and the Enlightenment, if not history itself.
David: Actually, I'll put it down to what a bad idea Parliamentary government is.
02 December 2008
Subscribe to:
Posts (Atom)