Felix Salmon

Schwab’s lies

Felix Salmon
Jan 14, 2011 15:42 UTC

Floyd Norris has a great column today taking Charles Schwab to task for its egregious behavior surrounding the ill-fated YieldPlus fund. Charles Schwab lied about the fund to the people who invested in it, and has not apologized for doing so: instead, the company just says that “we regret that fund shareholders lost money in YieldPlus.” Yeah, I’ll bet they do.

YieldPlus was Schwab’s largest bond fund—its real flagship in the fixed-income space, designed and marketed as a cash alternative. The company trumpeted its short duration and low weighted average maturity, but in fact the fund was investing in very long-dated instruments indeed: on Sallie Mae bond, for instance, didn’t mature until October 2021.

Schwab used two tricks to avoid telling investors just how long-dated the YieldPlus bond portfolio really was. One of them was to quietly stop talking about weighted average maturity, and start talking instead about modified duration. This is from the SEC complaint. (Warning: that link will take you to a 6MB PDF, created by scanning pieces of paper. Come on, SEC, you can do better than that.)

A fund’s weighted average maturity (“WAM”) is a measurement of the average length of time until the underlying bonds in a portfolio mature. WAM can be used by investors to evaluate the riskiness of a product; among similar funds, those with a longer WAM generally involve more risk. A fund’s duration is different than WAM; duration is a mathematical measure of a fund’s sensitivity to interest rate risk, but is not a measurement of time. For the relevant period, the YieldPlus Fund’s duration was a lower number than its WAM.

Between February 2006 and September 2007, in some communications with investors, Schwab substituted the Fund’s duration for its WAM, in some instances without noting the change. The resulting understatement appeared in sales and marketing materials and one Commission filing…

In early 2006, the YieldPlus Fund’s WAM increased significantly… CSIM and CS&Co. then listed the Fund’s duration in place of its WAM in the sales materials, including tables that listed statistics for all Schwab’s funds and, internal daily reports. Schwab did not replace WAM with duration for any other fund.

In some communications, CS&Co. and CSIM noted the replacement with a footnote indicating that duration, not WAM, was listed. However, in tables on the Schwab.com website, one Commission filing, and two issues of On Investing magazine, CS&Co. and CSIM did not include the footnote. As a result, for eighteen months, the website indicated that the average maturity of the Fund’s bonds was six months when the Fund’s WAM actually ranged from at least 1.3 to 2.2 years.

This is even more egregious than it sounds. It’s unclear to me exactly how Schwab was calculating the duration of the YieldPlus fund, but it looks as though the company went to great lengths to pick a calculation which measured only interest rate risk, rather than a more conventional measure. Bond duration is, as the excellent Wikipedia article on the subject says, the weighted average of the times until a bond’s fixed cash flows are received. In the case of floating-rate bonds it’s hard to measure duration accurately, since you don’t know how big certain future cash flows are going to be. But if you have bonds in your portfolio which don’t fully mature until 2021, you know that some weight has to be given to long-dated flows.

Duration is one way of measuring interest rate risk, but that’s not all duration does. It also gives one indication of market risk—the risk that the value of a debt instrument will fall. If it’s very unlikely that Sallie Mae will default in the next six months, then there’s a limit to how much the price of a Sallie Mae bond maturing in six months’ time can fall. On the other hand, if the bond just floats to a different interest rate in six months, then the price can fall a lot, since there’s still substantial default risk in the years ahead. So if you use a measure of duration which ignores maturity dates and just looks at interest-rate risk, then you’re leaving out a crucial piece of information—especially if at the same time you’re no longer giving out numbers for weighted average maturity.

But it gets worse than that. Norris explains that when Schwab did give out numbers for weighted average maturity, it was lying:

If an investor was worried about maturity risk, the 2007 annual report was reassuring. It said that on Aug. 31 of that year, more than 60 percent of the fund’s assets had maturities of six months or less. In the glossary at the back of that annual report, maturity was defined to mean just what it really means: “The date a debt security is scheduled to be ‘retired’ and its principal returned to the bondholder.”

But that was not what Schwab really meant. At the beginning of the list of investments held by the fund, it said that the maturity date shown for adjustable rate securities was “the next interest rate change date.”

There is no reason to do this, except that you want to deliberately mislead your investors into thinking that the fund is safer than it really is. Maturity is a measure of maturity, not a measure of the amount of time until a coupon resets. All this trickery with duration and maturity was designed with one purpose in mind: to lie to investors. Which Schwab did in other ways, too:

While the YieldPlus Fund’s NAV declined, CSIM, CS&Co., Merk, and Daifotis held conference calls, issued written materials, and had other communications with investors that contained a number of material misstatements and omissions concerning the fund. For example, in two conference calls, Daifotis made false and misleading statements that the fund was experiencing “very, very, very slight” and “minimal” investor redemptions. In fact, Daifotis knew that YieldPlus had experienced more than $1.2 billion in redemptions during the two weeks prior to the calls, which caused YieldPlus to sell more than $2.1 billion of its securities. Similarly, Merk authored, reviewed and approved misleading statements about the fund, such as a false claim that the fund had a “short maturity structure” that “mitigated much of the price erosion” experienced by its peers.

As Norris says, this kind of behavior flies in the face of Schwab’s ostensible commitment to be “on the side of the investor”: it’s not only illegal, as the company’s $118 million settlement with the SEC suggests, but it also destroys the main selling point which Schwab uses to differentiate itself from other brokers.

Schwab clients should I think pay great attention to the fact that the company has expressed precious little remorse over its actions surrounding YieldPlus, and indeed has started pointing fingers instead at “the investment banks that created mortgage-backed securities and the ratings agencies that legitimized them with triple-A ratings”. That says to me that Schwab’s culture hasn’t changed, and that investors in its funds can’t trust what they’re told. The company might be a good discount brokerage. But it’s not a respectable fund manager.


Your other commenters are right Felix.

A) What Schwab was almost certainly providing is more accurately described as “effective duration”. As another pointed out, the “weighted avg time to maturity” calculation is Macaulay Duration, which has other useful applications (liability matching), but isn’t of much help in cases like this.

B) The maturity issue is relevant only insofar as it helps point in the direction of the credit risk that was otherwise being obscured.

I understand why you interpreted and said “it also gives one indication of market risk”, but that’s just not so. Duration is not designed to bake in basis and credit risk. It has limitations, plain and simple. The only way to avoid the problem of duration not telling the whole story is to use other measures and metrics. Would a nominal maturity or weighted avg live number been helpful in this case? Absolutely.

Even that wouldn’t have told the whole story, though. In theory I could have constructed a portfolio with just about identical durations and WALs to the Schwab fund, but entirely comprising agency-backed MBS. Its risk would have been entirely different, and not because those measures were wrong. I can show you portfolios with those constructions–i.e. relatively short duration, long nominal maturity–that had relatively decent performance in 2008. Their credit/default risks were different; the Schwab fund had significant default risk underneath. Full Stop.

C) Does any of that mean Schwab’s managers WEREN’T trying to deceive clients? Absolutely not. Whether the ratings reflected the portfolio’s credit risk in real time,, the Schwab managers knew what was in those portfolios and how risky the collateral was. Those deceptions, to the degree that they existed, were the product of omission and misdirection, not the duration and maturity measures themselves.

Posted by fixedincome | Report as abusive


Felix Salmon
Jan 14, 2011 05:49 UTC

Why you should never, ever use two spaces after a period — Slate

IBM builds a supercomputer to play Jeopardy — AP

Pharo to offer shares denominated in renminbi — FT

How much would you bid for a piece of paper with a chap’s name on it? — eBay

NJ gov Christie’s idoitic ‘bankruptcy’ comments derail bond issue — Bloomberg

Joe Nocera is moving to the NY Times op-ed page — Forbes

Mark your diaries for April 30: it’s the New Amsterdam Bicycle Show! — NABS

Amy Chua says the Wall Street Journal misrepresented her book with its excerpt — SF Gate

J.P. Morgan Chase to end services for diplomats — WaPo

“Every day, 50,000 new men are trying to get naked,” he says. “What we’re doing is selling the naked men” — Fast Company


Actually, I have seen technical writing editors frothing about the double space. However, as far as I can tell, Word automatically adjusts the double space width to the same width as a single space in its proportional types so it doesn’t appear to significantly impact the visual appearance of a document.

So, it appears that technology has solved the problem that it caused for us that learned on old typewriters.

Posted by ErnieD | Report as abusive

The US won’t default, even if the debt ceiling stays

Felix Salmon
Jan 13, 2011 21:22 UTC

Greg Ip makes a very important point today, which I haven’t seen made anywhere else*: even if the US debt ceiling isn’t lifted, that doesn’t mean the government will default.

In any given month, the government’s income dwarfs its debt-service obligations, which means that the government could simply pay all interest on Treasury bonds out of its cashflow. Greg hasn’t run the numbers on principal maturities, but I’m pretty sure that they too could be covered out of cash receipts—and when that happened, of course, the total debt outstanding would go down, and we wouldn’t be bumping up against the ceiling any more.

The point here is that the government has enormous expenditures every month, and debt service constitutes an important yet small part of them. If the debt ceiling weren’t raised, it stands to reason that just about any other form of government spending would get cut before Tim Geithner dreamed of defaulting on risk-free bonds.

Some of those spending cuts could be implemented almost invisibly. For instance, Social Security runs a surplus for the time being; it invests that money in special non-marketable Treasury securities, which count as Treasury debt. If the Social Security trust fund accepted instead just some kind of promise of a top-up at a later date, that could save billions of dollars right there.

Beyond that, large defense contractors aren’t going to stop working for the government just because they’re late in being paid; neither are doctors, hospitals or most of the rest of the healthcare industry.

But maybe the smartest thing for Geithner to do would simply be to stop paying the salaries of members of Congress and their staffs. It probably wouldn’t take long, in that event, for Congress to vote Obama the debt-ceiling raise he needs.

The bigger picture here is that the US government, like any other company or individual, has enormous freedom when it comes to which creditors it chooses to pay when. Just like GM had every right to privilege some creditors over others, even when those creditors were legally pari passu, the US government can do exactly the same thing. And there’s no way that this administration, or any other that I can think of, would choose to cut debt service given that they have every choice in the matter.

*Update: Which doesn’t mean the point wasn’t made earlier elsewhere, of course. Stan Collender made it in his Roll Call column, which he posted on his blog here. Apologies, Stan, for missing it.


RobSterling makes a decent point: If the main consequence of a failure to raise the debt ceiling is that the administration has to make an ongoing series of painful cuts to keep paying its bills, all the incentives for Republican members of Congress will still be to vote against raising the debt ceiling.

GOP reps want to get the credit for shrinking government without copping the criticism for cutting specific programmes. This situation would give them exactly that situation, with the administration facing an escalating series of Sophie’s Choice moments on spending cuts–decisions for which it, and it alone, would be seen as responsible–until it gives in to whatever concessions the GOP would demand in return for finally raising the debt ceiling.

I don’t see this administration responding by just targeting programmes that Republicans like: they have taken such pains to appear centrist and reasonable, why throw that all away with a set of cuts that would look vindictive?

It’s much more likely that if we got to this point, the first stuff to go once accounting gimmicks, phantom surpluses etc are used up would be stuff that appeals more to Democrats than Republicans. That way independents would see Obama as being ‘fair-minded’.

So I guess we might not have a default, but I think the period after the debt limit runs out could be more drawn-out and painful than people are anticipating.

Posted by vanityvehicle | Report as abusive

Algorithmic trading and market-structure tail risks

Felix Salmon
Jan 13, 2011 17:50 UTC

I’m on Fresh Air today, talking about the Wired article on algorithmic trading which I wrote with Jon Stokes. Norm Cimon, who wrote a very interesting paper on the limits of financial databases, read it, and emailed with an important point:

The marketplace is, in mathematical parlance, a non-linear iterated discrete dynamical system. Given the introduction of all that computing power, however, it’s one that can now crash and burn (really a transition to a different portion of the state space) in an instant.

There’s a real lack of understanding about such systems. They are not predictable because they require perfect knowledge of the state at any given time and that is, of course, not possible. Trajectories for systems like this can transition instantaneously, not because there’s something wrong with them but because they’ve been wired up to operate that way. There’s been very little discussion which isn’t all that surprising. I don’t think there’s much contact between those trained in dynamical systems theory and Wall Street. That’s too bad. The quants would have been wise to spend a little less time on probability and statistics, and a little more on nonlinear dynamics.

On second thought, it wouldn’t have done them much good, since it’s really a qualitative as opposed to a quantitative science. It will give you a great idea of what might potentially happen under various system forcings, and no predictive power to say when.

The University of Pennsylvania’s Michael Kearns is briefly quoted in the piece, but I actually talked to him for quite some time about these issues. He’s a real expert on these matters, who’s done a lot of work on Wall Street, and he’s very worried about issues of market structure. This is probably as good a place as any to put a few chunks of the interview which fell onto the cutting-room floor, as it were.

Here’s some of what Kearns told me. It’s all pretty much verbatim from our conversation, and more or less self-explanatory. If you like geeking out about market-structure tail risks, there’s lots of interesting stuff here.

There’s a growing understanding and belief on Wall St, especially intraday, that there’s a strong game-theoretic aspect to the market: the performance of a given strategy depends on what other strategies are trading in the market at the same time. My payoff is a function of what I do and what the other players in the game do.

The concepts of strategic interaction are important. Game theory is hard to understand even in simple cases. And now strategies are adaptive, so it’s especially difficult. I’m not frightened by it, but it’s a legitimate concern. People who build good machine learning algos have a fair amount of knowledge about what they’re doing, but they’re still adaptive. It’s natural to have the model retrain itself each night based on each day’s data. Humans might not know what incremental modifications today’s data introduced in this adaptive process. If everybody’s doing this, you can easily imagine various effects. That could be a bad thing from a stability standpoint in the markets.

Our financial markets have become a largely automated adaptive dynamical system, with feedback. Furthermore, a dynamical adaptive system with feedback is challenging ordinarily, but this one is also strategic. A flight controller is dynamical, but not strategic. Add an adversarial environment, and there’s no science I’m aware of that’s up to the task of understanding the potential implications of that system.

The quant meltdown demonstrated the unexpectedly high correlation between quant strategies who believed they were doing different stuff from each other. Under ordinary circumstances these correlations weren’t so apparent. When we decide to exit our positions at the same time, we’re going to painfully realize our correlation.

There’s some echo of that in adaptive algos: if we’re all using somewhat similar adaptive algos on the same data, then the data itself correlates us. If we were forbidden from backtesting, that would help: that’s the way it was in the old days. Pre computers, there was no backtesting. You had a hunch, and you played it. Now, with shared data, this provides an opportunity to correlate our behavior without realizing it.

At any given moment, there aren’t that many ideas which make money. So the firms which have survived — are going to be the ones that discovered those ideas. And they’re doing similar things. The quant meltdown revealed the high correlation between quant trading strategies, and the flash crash was almost more of a microstructure. Something got slowed down in some particular exchange, causing a flight of liquidity from that exchange.

The quant meltldown was more macroscopic, the flash crash showed how microscopic events, with automated algo trading, could be amplified in unexpected ways.

The flash crash was more or less survived, for the most part. The quant meltdown wasn’t, for most groups. These concerns are not going to go away, and there aren’t obvious fixes to them.

What worries me most is the idea that any simple proposals, that are only minorly disruptive, will immunize us. You can prevent the flash crash, but it’ll be something different next time.

Kearns has one idea which he thinks might help in understanding these risks: a public-domain stock-market simulator.

If I was going to make one high-level recommendation, I’d say that research needs to be done. Who builds sophisticated market simulators? Groups that want to use them for profit. I feel that recent events have shown us that we need some kind of science for simulating the vulnerabilities of the market structure that we have.

You can imagine trying to build an ambitious, reasonably faithful simulator of our current markets. You’d have high-frequency algos, shorter-term stuff, dark pools, multiple exchanges, etc. A giant sandbox. Then you’d ask questions: in some simulation, what happens if, for instance, one exchange slows down.

If you do a simulation and you try some perturbation or stress, and it tells you that a disaster happens, then it’s worth thinking hard about our current markets. But if you don’t find a disaster, that’s not reassurance that some other disaster won’t happen.

I’m proposing a quant version of the stress tests that were proposed for banks.

There’s no way, in some sandbox, of testing what all the effects might be. A car company, before they roll out a product, have a lab environment where they put it through tests. And in reality problems which weren’t tested for get discovered. We’d be much better off with a simulator.

We have no such lab for our financial markets. This strikes me as a little off.

People on Wall St think about simulation, but not for catastrophe prediction, just for their own trading purposes.

For some kinds of research, we should be thinking about it as a nation. Almost anything would be better than what we have now, which is nothing, beyond field experiments.

History can only give us insight into the particular catastrophes that have already occurred. We haven’t come near exploring the range of catastrophes which could befall us. There are things we could do to make it better. I applaud looking at the tape, but too much focus on that will give us false reassurance.


y2kurtus, you forget… The Big Boys aren’t allowed to lose money. That threatens the system. Only individual investors are allowed to lose money.

Posted by TFF | Report as abusive

The business lobby takes on healthcare reform

Felix Salmon
Jan 13, 2011 14:54 UTC

What business, big or small, doesn’t like a government bailout? That’s what the Obama healthcare bill is, in oversimplified terms, from the point of view of employers: up until now they’ve been struggling with soaring healthcare costs, and now the government is stepping in to relieve some of that burden.

David Wessel finds it necessary today to explain this at some length, thanks to the fact that the US Chamber of Commerce and the National Federation of Independent Business both want to repeal the bill.

At the National Business Group on Health, a collection of nearly 300 big employers, President Helen Darling, a former corporate-benefit administrator and Republican Senate staffer, says about executives who call for repeal: “If they really understood it, they wouldn’t.”

Wessel explains that the bill will help minimize the implicit subsidies that insured employees pay to the uninsured; will make it much easier for small businesses to shop for healthcare; will end Cobra obligations; and might even slow the pace of healthcare cost inflation.

What he doesn’t explain is how the US Chamber of Commerce and the National Federation of Independent Business became party-political hack machines, lobbying for whatever’s good for Republicans politically rather than whatever’s good for businesses on a policy basis. Is this something new, something related to the increasingly-partisan nature of Washington? Or were they ever thus?

It makes sense that there should be organizations which aggregate and represent the view of businesses in the economy — but maybe such organizations simply can’t be based in Washington and also be intellectually honest. After all, when politics and economics meet, politics always wins.


Executives don’t always lobby for what is best for their company, they lobby for what is best for themselves


Posted by lambertstrether | Report as abusive


Felix Salmon
Jan 13, 2011 06:22 UTC

The ultimate Steve Jobs fantasy — a truly buttonless device – looks like it’s coming true at Apple — BGR

Charity Navigator’s guide to where the Haiti money went includes only 1 of the 5 telethon recipients — Charity Navigator

22% of Americans think we have double-digit inflation, and 18% believe former BP chief Tony Hayward is the UK prime minister — Pew (PDF)

“Treating teachers like professionals” cuts both ways. If you move to defined-contribution pensions, then be sure to jack up pay first — Public Sector Inc

“South Sudan will be the 195th country on the official State Department list, and the 49th sub-Saharan republic” — DLC

On the Richness of the Rich — DeLong

Now that was unexpected: AP and Shepard Fairey “to collaborate on a series of images that Fairey will create based on AP photographs” — AP


Teachers are decently paid, at least in my state. It isn’t an easy job by any stretch of the imagination — a good teacher will put in 2000 hours between September and June, and some more over the summer as well. But the pay is respectable.

If the goal of switching to a defined-contribution plan is cost-certainty (rather than cost savings), then schools should adopt a contribution match that follows the college standard rather than the private sector model that emphasizes a high base pay with fewer benefits.

Posted by TFF | Report as abusive

Dealing with Britain’s overpaid bankers

Felix Salmon
Jan 13, 2011 05:57 UTC

Bagehot has a very odd column about Britain’s overpaid bankers. Part of it is spot on:

One shorthand description for the New Labour boom years is: Gordon Brown let a deregulated City rip, then used the tax revenues to fund a dramatic expansion of the state.

He’s quite right about this. If a government starts seeing tax revenues from banks rise sharply, it should worry: that’s a sign of a dangerous financial bubble. The exception to that rule, of course, is when a government deliberately tries to cut the financial sector down to a more sensible and sustainable size by taxing it more.

But that’s not the way that Bagehot sees it. For him, the question of whether bankers are paid too much is exactly the same as the question of whether, if they were paid less, they would move to some other country.

That’s silly. An overpaid banker in Hong Kong or Sao Paulo is still an overpaid banker. And the UK must make its own determination as to whether or not it wants to be home to a large contingent of overpaid bankers.

Bagehot might be right that if London’s contingent of overpaid bankers were to shrink, then its financial-sector tax revenues would probably shrink as well. But if you’re addicted to the fiscal crack cocaine that is City taxes, that’s a reason to give up those taxes — it’s not a reason to keep on going back for an extra fix, even after bitter experience has taught you how damaging your addiction will prove to be in the long run.

Bagehot writes:

Yes, Britain’s economy is dangerously dependent on revenues from the financial sector. It is equally worrying that the City of London offers one of the few claims Britain has left to global prominence. But rebalancing the economy by whacking the City with a mallet is surely rather an imperfect solution: is it too much to hope that Britain might achieve a better balance by building up other sectors of the economy or regions of the country?

The problem here is that so long as the City remains dominant in the UK economy, other sectors of the economy and regions of the country will never be able to attract the smart labor it desperately needs.

The answer to the question of whether bankers are overpaid can be seen not in a hypothetical exodus of bond traders to Frankfurt, but rather in the fact that far too many of Britain’s brightest graduates end up in the financial sector, instead of building up the rest of the economy. So long as the Pied Pipers of the City keep on playing their siren tune and luring the cream of Britain’s future to a dense and dangerous square mile of London, the rest of the country, bereft of its native talent, is certain to continue to underperform.


The Coalition has little incentive to cap bonuses as half of it comes back in taxes. Their only concern is that the issue has become a political stick with which Labour can beat them. Labour could evolve an alternative approach now, if they could see beyond the immediate benefits of beating up the Coalition using bonuses as the media-wielding stick. A State Savings Bank (RBS or perhaps Lloyds TSB with its branches and without a significant trading division?) would guarantee depositors and not indulge in hedge fund and derivatives trading. It could genuinely invest in UK-based small business and support first-time buyers! Depositors in all other banks would be told that there would never be any government guarantee of their deposits (no more bail-outs). Other banks could then do what they wanted (no more vain attempts to restrict their operations as all State prohibition policies are always doomed to fail) … but all the risk would be theirs not the taxpayers.

Posted by F1ddler | Report as abusive

The behavioral case for the debt ceiling

Felix Salmon
Jan 12, 2011 22:56 UTC

John Carney takes a stab at answering my question about why we have a debt ceiling, saying that it “helps raise public awareness about the costs of government”:

Lawmakers must go on record as approving an increase in the debt limit in order to enable the government to borrow to fund the spending the lawmakers have approved. They must confront, in a very public manner, the costs of the programs they have enacted.

I’m not at all sure this is the actual reason why we have a debt ceiling; at best it’s a behavioral reason not to abolish it. But it’s not a good reason.

Let’s take a personal-finance analogy here. A credit card company comes along and offers you a card with a $200,000 credit limit—much more than you should ever borrow. But you know that if the money is there for the spending, you’ll be more likely to spend it. So you phone up the credit card company and do a deal with them. You set the limit very low—$1,000, say—and then, any time you want to go over that limit, you have to phone up the credit card company and get them to raise it.

You know in advance that the credit card company will say yes: after all, they’ve already told you that as far as they’re concerned they’re happy to lend you $200,000. But being forced to phone them up and ask for a credit-line increase helps to drive home the consequences of your spending decisions, in a way that simply whipping your card out at the Apple Store doesn’t.

So far so good. But what if you were using your card not only to buy iPads, but also to make rent? And what if there were possible glitches with the system whereby you phoned up and asked for a credit-limit increase, so that you couldn’t be sure it would always work? And what if a single late rent payment could be catastrophic, ending up with you thrown out of your home? At that point, your clever system would stop seeming so clever, and start seeming downright risky.

And that’s the problem with the debt ceiling. It might have interesting and possibly even beneficial behavioral second-order effects, although there’s precious little evidence for that. But getting those beneficial effects means playing with fiscal high explosives, which run the risk of blowing up in the economy’s face and causing major damage. It simply isn’t worth it.


@TFF, thanks for pointing that out!

I think the Fed’s ability and awareness to fight deflation makes all the difference in the world.

In massive deflation asset prices collapse as they are sold at inappropriate fire-sale prices to raise emergency funds. Economics essentially breaks down and great businesses are destroyed as the line between illiquid and insolvent vanishes.

To me it is perfectly reasonable to be fiscally tight and monetarily loose, but that is by no means my original idea. That is the old and standard IMF solution, completed successfully in countless countries around the world. IMF or not, that seems a standard path in the end. Most recently that was the basic path in places like Argentina and Iceland, which have gone from total failure to stability, growth and jobs.

In fact fiscal tightness and monetary looseness is such a normal endgame that the only question may be when it happens and how many years and how much growth are lost in the interlude.

Posted by DanHess | Report as abusive

Algo-powered curation

Felix Salmon
Jan 12, 2011 21:17 UTC

Paul Kedrosky has a great little essay today on the way that curation is set to be the new search, since the utility of the old search is steadily declining.

Google’s ranking algorithm, like any trading algorithm, has lost its alpha. It no longer has lists to draw and, on its own, it no longer generates the same outperformance — in part because it is, for practical purposes, reverse-engineered, well-understood and operating in an adaptive content landscape.

Kedrosky rightly points to Twitter as one powerful source of curation, but there are and will be others:

The re-rise of curation is partly about crowd curation — not one people, but lots of people, whether consciously (lists, etc.) or unconsciously (tweets, etc) — and partly about hand curation (JetSetter, etc.). We are going to increasingly see nichey services that sell curation as a primary feature, with the primary advantage of being mostly unsullied by content farms, SEO spam, and nonsensical Q&A sites intended to create low-rent versions of Borges’ Library of Babylon.

I’m enthusiastic about this to the point of being downright conflicted: I’m involved in putting together one such nichey service myself (on which more anon), and can’t wait for a whole ecosystem of such things to arrive.

Up until now, publishers have tended to be suspicious of anything which is built on the idea of sending people away to algorithmically-generated corners of the internet which aren’t all carefully vetted. But if you can get over that conceptual hurdle, there’s a very exciting world out there which has barely been explored at all—a world of vibrant websites whose content is powered by algorithms but still produced, curated, and edited by very smart humans. Techmeme is a good example; there will be many, many more.


Hi Felix,

You’re right algorithmic curation is definitely an interesting area at the moment especially with real time content.

If you’ve not already heard about DataSift it sounds like it may be something of real interest to you, DataSift is a real time content curation engine for the social web. It has multiple input sources and a range of different augmentation options and you have control of how the data is curated through our CSDL (Curated Stream Definition Langage).

If you are interested then do sign up to our Alpha programme. http://datasift.net

Many thanks

Posted by DataSift | Report as abusive