Information Processing

Just another weblog

Archive for June 2005

FX update

with 5 comments

According to BIS figures published this week global central bank currency reserve diversification has been larger and faster than previously thought. Stories of central banks divesting their dollar holdings have been common in the past two years. However, hard evidence is difficult to come by as official data is only available on a sporadic basis in the form of the BIS and IMF annual report. The latest BIS annual report published this week, therefore, provides an important benchmark to the extent and pace of diversification. In a break from the past, this year the BIS published a series of the dollar’s share of global FX reserves in constant (2003) exchange rates. This controls for the variations in exchange rates that can depress or exaggerate a currency’s share in reserves. For example the weakness of the US dollar in the past three years would tend to deflate the dollar’s share of reserves even if the actual holdings of dollar reserves were constant.

Surprisingly, the BIS numbers are somewhat different to those published in the IMF annual report in September 2004. At the end of June 2004, the BIS estimate that 65.5% of reserves were held in US dollars. While this is not significantly different from the IMF reading of 63.8% at the end of 2003, the pace of reduction in dollar holdings is far quicker in the BIS data set. The share has fallen from 77% in 2001 (see chart for more details). The BIS also shows that the euro has been the main beneficiary of this change, with the share of euro reserves more than doubling since 1999 to 23.6%. (Source: State Street, via our correspondent)

At present the main currencies are not far from their post BW I averages in real effective terms. The NZD, GBP, TWD, SEK and SGD were the only currencies that were more than 15% off their respective averages:

Written by infoproc

June 30, 2005 at 5:41 am

Posted in globalization

Google payments

with 2 comments

Rumors have been around for some time that Google will roll out a payment system. Company representatives have said it won’t be a person to person payment system like PayPal. I suspect it will be a micropayment system that lets users pay for content out of their Google Wallet. This is a crucial component of the Internet ecosystem that has been needed for some time, and Google is one of the few entities with enough clout, eyeballs and technical capability to roll it out. It will enable Web publishers large and small to sell content, going beyond the current Google AdWords model which allows ad sales. This would allow publishers ranging from the very big (Slate, Salon, New Yorker, Time, etc.) to the very small (individual bloggers) to monetize both new and old content.

In a related development, Google is readying a platform that allows people to upload video and make it available for distribution. A natural step will be for video auteurs to charge a fee for each viewing, or for Google to split fees with content providers for footage from news, sports, movies, etc. Google has already reportedly made a lot of progress in video search.

While I’m impressed with Google’s string of innovations (images, Gmail, maps, etc.) I still think the stock is overpriced at $300 🙂

Written by infoproc

June 28, 2005 at 5:33 am

Posted in Uncategorized

AI update

with 3 comments

Two tidbits from the inexorable advance of artificial intelligence.

1) Two amateurs from New Hampshire (a database administrator and a soccer coach) won a recent international “freestyle” chess tournament, which included several grandmasters. Freestyle chess is team competition, including both humans and computers. The winning team (ZackS, anonymous throughout the tournament) used ordinary PCs and commercial chess software. Nevertheless, their play was so spectacular that many suspected the presence of Gary Kasparov!

The other untitled team, ZackS, is a dark horse. The identity of the people behind this team, and the method they are using, will be revealed after the tournament is over. Everybody assumes that there are one or more GMs working together with the team captain. The rumour was that Garry Kasparov was producing the extraordinary chess displayed by ZackS, but we can confirm that on the weekend of the quarter-finals Kasparov was most certainly otherwise engaged.

The standard of play is very high, possibly the highest ever seen in chess at these time controls. One would scarcely expect a human player, even the best in the world, to be able to face the precision and the strategic depth of some of the participants in this event.

2) An AI program has matched the average human performance on the verbal analogies portion of the SAT. (You know, “fish is to sea as monkey is to …?”) This is far short of passing the Turing test, but still an impressive feat of extracting relations from computer analysis of a terabyte of text. The program was written by Peter D. Turney’s Interactive Information Group, Institute for Information Technology of the National Research Council Canada.

Written by infoproc

June 26, 2005 at 6:47 am

Posted in Uncategorized

How you got here

leave a comment »

Below are the top keyword searches that led people to this blog in the past 24 hours. Apparently my post on the time travel movie Primer is popular. Other topics of interest: China, hedge funds, globalization, financial bubbles and the occasional wormhole…

Num Perc. Search Term
10 25.00% primer movie
3 7.50% china climbing
2 5.00% single-digit millionaires
2 5.00% why not hyperinflation
2 5.00% primer/movie
1 2.50% emerging markets debt processing
1 2.50% analyzing hedge fund returns
1 2.50% information processing in brain
1 2.50% advantages and disadvantages of globalization
1 2.50% ltcm today
1 2.50% horvitz cleveland
1 2.50% daniel kahneman and amos tversky
1 2.50% jeremy grantham 2005 interview
1 2.50% john d gartner the hypomanic hedge
1 2.50% cds implied volatility
1 2.50% asness bubble logic
1 2.50% tennessee candidate issues
1 2.50% china-japan relations
1 2.50% all about wormholes
1 2.50% cdx index
1 2.50% china explaining high savings
1 2.50% price to rent ratios
1 2.50% sony vs samsung tv
1 2.50% global real interest rates
1 2.50% why long bonds
1 2.50% price to rent ratio

Written by infoproc

June 23, 2005 at 4:02 pm

Posted in Uncategorized

Greetings from startup land

with one comment

Sorry for the lack of posts recently. I’m in the bay area right now, setting up our new offices. Commercial real estate is still a bargain here – we looked at a number of places at around $1 per sq ft, full service. For example, I happened to see a nice 2000 sq ft live-work loft near the Berkeley-Oakland border for only $2k per month. Why buy an $800k house when you can rent a nice space like that? Certainly says something about price to rent ratios here 🙂

We ended up renting a place between downtown Oakland and lake Merritt (not far from Chinatown). It’s an up and coming area with lots of interesting restaurants. We’ve filled our 2000 sq ft offices with IKEA office furniture and Dell computers (very happy to be rid of that horrible U-Haul truck). My office here is now nicer than at the university or at home… The last tenants left us a closet filled with switches and some old servers, so they were pretty sophisticated themselves. I guess every small business these days has some network plumbing in the walls.

We’re pretty happy with the team we’ve hired. We found the majority of our guys via Stanford and Berkeley, but struck out this time at Caltech, my alma mater!

Now it’s into stealth mode… look for our beta release near the end of the year 🙂

Written by infoproc

June 22, 2005 at 6:29 am

Posted in Uncategorized

Economist on global housing bubble

with 2 comments

Informed readers are already familiar with the situation, but the Economist as usual does a wonderful job of summarizing. In the first figure, note the oscillation of price/rent about the long-term average. In the second figure, you see a possible decades-long unwinding of the bubble, as in Japan.

Robert Shiller (quoted in the article) has called this the biggest bubble of all time, and the Economist agrees. In addition, his research shows only a tiny (.4% per year) real return on US house prices over the last century, contrary to conventional wisdom.

The most compelling evidence that home prices are over-valued in many countries is the diverging relationship between house prices and rents. The ratio of prices to rents is a sort of price/earnings ratio for the housing market. Just as the price of a share should equal the discounted present value of future dividends, so the price of a house should reflect the future benefits of ownership, either as rental income for an investor or the rent saved by an owner-occupier.

Calculations by The Economist show that house prices have hit record levels in relation to rents in America, Britain, Australia, New Zealand, France, Spain, the Netherlands, Ireland and Belgium. This suggests that homes are even more over-valued than at previous peaks, from which prices typically fell in real terms. House prices are also at record levels in relation to incomes in these nine countries.

…To bring the ratio of prices to rents back to some sort of fair value, either rents must rise sharply or prices must fall. After many previous house-price booms most of the adjustment came through inflation pushing up rents and incomes, while home prices stayed broadly flat. But today, with inflation much lower, a similar process would take years. For example, if rents rise by an annual 2.5%, house prices would need to remain flat for 12 years to bring America’s ratio of house prices to rents back to its long-term norm. Elsewhere it would take even longer. It seems more likely, then, that prices will fall.

A common objection to this analysis is that low interest rates make buying a home cheaper and so justify higher prices in relation to rents. But this argument is incorrectly based on nominal, not real, interest rates and so ignores the impact of inflation in eroding the real burden of mortgage debt. If real interest rates are permanently lower, this could indeed justify higher prices in relation to rents or income. For example, real rates in Ireland and Spain were reduced significantly by these countries’ membership of Europe’s single currency—though not by enough to explain all of the surge in house prices. But in America and Britain, real after-tax interest rates are not especially low by historical standards.

…Another mantra of housing bulls in America is that national average house prices have never fallen for a full year since modern statistics began. Yet outside America, many countries have at some time experienced a drop in average house prices, such as Britain and Sweden in the early 1990s and Japan over the past decade. So why should America be immune? Alan Greenspan, chairman of America’s Federal Reserve, accepts that there are some local bubbles, but dismisses the idea of a national housing bubble that could harm the whole economy if it bursts. America has in the past seen sharp regional price declines, for example in Boston, Manhattan and San Francisco in the early 1990s. This time, with prices looking overvalued in more states than ever in the past, average American prices may well fall for the first time since the Great Depression.

But even if prices in America do dip, insist the optimists, they will quickly resume their rising trend, because real house prices always rise strongly in the long term. Robert Shiller, a Yale economist, who has just updated his book “Irrational Exuberance” (first published on the eve of the stockmarket collapse in 2000), disagrees. He estimates that house prices in America rose by an annual average of only 0.4% in real terms between 1890 and 2004. And if the current boom is stripped out of the figures, along with the period after the second world war when the government offered subsidies for returning soldiers, artificially inflating prices, real house prices have been flat or falling most of the time. Another sobering warning is that after British house prices fell in the early 1990s, it took at least a decade before they returned to their previous peak, after adjusting for inflation.

Written by infoproc

June 18, 2005 at 3:39 am

Posted in globalization

Asian-Americans hurt by affirmative action

with 13 comments

From Gene Expression, results of a new Princeton study showing that if affirmative action were eliminated at elite universities, 80% of the previously reserved slots would go to Asian Americans. I predict little or no protest from model-minority Asians over this. The paper can be found here, but you need a subscription to Social Science Quarterly to read it.

The researchers’ results seem to be in agreement with what happened at Berkeley after UC was forced to drop affirmative action – the main effect was a drop in the numbers of black and hispanic students, a big increase in the number of Asians and little effect on the white population.

Why wouldn’t white and Asian-American applicants benefit equally if admission were purely by merit? It sounds suspiciously like the quota system imposed on Jews early in the 20th century. Previous research by these authors showed that being Asian was statistically equivalent to a penalty of 50 points on SAT score. (Probably due to preference awarded to “legacies”, who are predominantly white.)

Disregarding race in college admissions would cause sharp drops in the number of black and Hispanic students at elite institutions, according to a new study by two researchers at Princeton University. The study, described in an article published in the June issue of Social Science Quarterly, also found that eliminating affirmative action would significantly raise the number of Asian-American students, while having little effect on white students.

If affirmative action were eliminated, the acceptance rates for black applicants would fall to 12.2 percent from 33.7 percent, while the acceptance rates for Hispanic applicants would drop to 12.9 percent from 26.8 percent, according to the study. Asian-American students would fill nearly 80 percent of the spaces not taken by black and Hispanic students, the researchers found, while the acceptance rate for white students would increase by less than 1 percent.

The researchers who conducted the study — Thomas J. Espenshade, a professor of sociology, and Chang Y. Chung, a statistical programmer at Princeton’s Office of Population Research — looked at the race, sex, SAT scores, and legacy status, among other characteristics, of more than 124,000 applicants to elite colleges.

Written by infoproc

June 17, 2005 at 4:18 am

Posted in Uncategorized