Felix Salmon

The cost of Bernanke’s failure-aversion

Felix Salmon
Sep 3, 2010 23:23 UTC

John Cassidy has very little patience for Ben Bernanke’s latest attempt, in front of the FCIC, to explain how Lehman Brothers was allowed to fail so catastrophically. Bernanke is now saying that Lehman was in such bad shape that it would have failed whether or not the Fed had stepped in to guarantee its debts; like Cassidy, I’m very suspicious of that argument, since a Fed guarantee would have stopped any bank run cold in its tracks.

So what does Bernanke mean when he says that “the view was that failure was essentially certain in either case”? My feeling is that Bernanke, along with Hank Paulson, had an unnecessarily binary idea of what exactly “failure” meant. They were faced with a choice between the chaotic collapse that we saw, on the one hand, and a much more orderly failure, on the other; and they utterly failed to grok how much worse the first option was than the second.

Bernanke has long said that the Treasury “did not have the authority to absorb billions of dollars of expected losses to facilitate Lehman’s acquisition by another firm” — but now it seems that he’s also saying something which demonstrates much weaker leadership. If we lose billions of dollars and Lehman still fails, goes the argument as I understand it, then we will have failed too. So we might as well just let Lehman fail on its own. Even if the consequences of that decision are orders of magnitude worse.

A leader will take a hit for the greater good. A profit-driven trader like Hank Paulson, not so much. As Cassidy puts it:

Many people from Lehman and Barclays suspect that the real barrier to the Barclays rescue wasn’t the legal niceties in London but a reluctance on the part of Bernanke and others—Treasury Secretary Hank Paulson in particular—to fill the gaping gap in Lehman’s balance sheet by providing a Bear-style loan from the Fed, which could have topped fifty billion dollars.

With hindsight, $50 billion would have been a very small price to pay for an orderly wind-down of Lehman Brothers. But Bernanke and Paulson, it seems, were too caught up in wanting to avoid “failure” to work that out.


TFF, the Fed publishes a report on the three Maiden lane vehicles. Currently the difference between the expected cash flows of the BSC investments and the loan given to buy the investments is around around -3.5 billion USD so the cost of the bailout assuming projections are correct is 2.5 billion to the tax payer.

Thats to be balanced by the nearly 4billion the taxpayers stands to make on those assets it acquired when it “backdoor bailed out” the investment banks when it closed out the AIGFP swaps.

OnTheTimes, you seriously suggesting that GS had zero exposure to LEH? that LEH going bankrupt had no affect on GS?

AABender1, BSC was different. There was a very defined exposure with BSC given that JP was taking the majority of the risk – the Fed’s risk was capped at 29 billion and that loan was collateralised relatively well as you can see at the relatively smal losses at a cost to you personally of currently 11 USD over two years. I am sure two Big Macs less in that time have not gone a wanting…. Weird no one hears any outrage about the bailouts of the automakers, only one of whom is even vaguely viable.

Posted by Danny_Black | Report as abusive

Will the FCIC report be a whitewash?

Felix Salmon
Sep 3, 2010 15:33 UTC

Barry Ritholtz was unimpressed with the way that the Financial Crisis Inquiry Commission was quite soft on Dick Fuld:

To think that Fuld’s brand of psychopathic revisionism was given a sympathetic hearing is deeply disturbing.

I haven’t written this before, but now I am compelled to: I now fear the FCIC report is going to be an ideological farce. The nightmare report scenario is a collection of false statements, half truths, misunderstandings, confirmation biases, and rhetorical nonsense.

Obviously, the proof of the pudding will be in the eating. But I think Barry is stretching too far, here. Public hearings are interesting and useful, because they give the commissioners the opportunity to ask important questions of the principals in the financial debacle. But there’s no reason why they should set the tone for the final report.

Specifically with regard to Fuld, the FCIC barely needs to call anybody or do anything beyond reading the Jenner Report in full. More resources went into putting together that report than are going into the entire FCIC apparatus, so it makes sense for the FCIC to essentially outsource Lehman to Jenner. Certainly asking questions of Dick Fuld in a public hearing isn’t going to tell them anything they don’t already know, so why bother.

It’s worth noting that Joe Cassano of AIG got off very lightly in front of the FCIC as well. And if there’s any unanimity at all about who belongs in the rogue’s gallery of Top Villains of the Financial Crisis, there’s unanimity over Cassano and Fuld. I doubt that either of them will receive any hint of exoneration in the final report. As a result, they don’t need to be stoned in public as well.

There are also time considerations for the commissioners: there’s little point in them spending a lot of time preparing to grill Fuld and Cassano, since there are relatively few contentious and open questions relating to what those two men did. Instead, their time is better spent working out exactly where the FCIC’s subpoena power can be put to best use in determining the causes of the crisis.

Remember, the FCIC isn’t some kind of court where bankers go to get prosecuted. And the key determinant of how the final report reads is not who said what in the FCIC’s public hearings, but rather who actually writes the report, and who, among the commissioners, has the most political clout when it comes to pushing their personal opinion. If it’s basically put together by the likes of Angelides, Born, and Holtz-Eakin, we’ll get something great. If, on the other hand, there’s a lot of input from people like Thomas and Wallison, we might end up with exactly what Ritholtz fears.

For the time being I’m hopeful, since Angelides is the chairman, he’s smart, and he’s on top of his brief. The commission does seem to have recently lost its lead writer, which is a little bit worrying, but nothing huge to worry about. So long as there’s someone there who can express complicated thoughts clearly, there’s hope yet for this report.


Well when it comes to writing a “collection of false statements, half truths, misunderstandings, confirmation biases, and rhetorical nonsense”, Ritholz is without peer. Virtually every single one of his claims are either complete BS or do not separate LEH from GS or MS or MER.

I think the reason the FCIC hearings are so unsatisfactory is that they seem determined to focus on the issue of capital or derivatives when the real issue for AIG and LEH was liquidity. For all the fuss repo 105 got, I have only seen one person comment online about the games they played with their liquidity pool and it is clear that LEH and BSC and others got into trouble because their liquidity dried up not because of the leverage and certainly not because of derivatives as most of the “problematic assets” were out and out real estate investments and leveraged loans.

Posted by Danny_Black | Report as abusive

Payrolls: Flat is the new up

Felix Salmon
Sep 3, 2010 13:14 UTC

Everybody was so nervous going in to this morning’s payrolls report that even though employment fell and unemployment rose, markets are looking extremely exuberant.

There’s definitely reason here for a small sigh of relief. The private sector added jobs in August — not enough to keep up with population growth, and not even enough to counteract the effects of census workers being laid off. But hey, the private-sector employment number was positive rather than negative, that’s gotta count for something.

The big picture here is, as Tony Fratto says, that the job market is basically flat. The official release generally has a pretty good take on things, and the language there is clear: the unemployment rate and the 14.9 million number of unemployed are “little changed in August”. Break it down by race or age, you still see “little change”. The labor force participation rate and the proportion of the population with jobs? “Essentially unchanged.” The number of people who want a job but didn’t actively look for one in the previous four weeks? Again, “little changed”.

The big number, total nonfarm payrolls, “was little changed”. Retail employment “was about unchanged over the month”. Elsewhere, employment “showed little change in August”. You get the picture.

Flat, then, is the new up — which only goes to demonstrate just how worried the markets are about a double-dip recession. The syllogism is easy. This payrolls report would never be good news in a growing economy; this payrolls report is good news; therefore, the economy isn’t growing. So don’t get too excited about bond yields rising today. We’re not remotely in full-bore recovery mode yet.


Dan, I generally don’t bother to respond to anybody who responds to my statements with political labels. Those who can’t think for themselves are not worthy of a response.

Other than that, I probably don’t disagree with him much. It is going to be a tough adjustment (for the US) before the gap between the US and Chinese labor markets is closed. This is part of why I don’t expect much GDP growth in the US for a while. Chinese wages are rising pretty rapidly, and there are some advantages to operating here, but it is a LARGE gap to close.

Nor do I disagree with you. Isn’t a pleasant situation, but the cure would be worse than the disease.

Posted by TFF | Report as abusive


Felix Salmon
Sep 3, 2010 07:10 UTC

Tony Blair, tagged — Yfrog

Blu-ray edition of “Withnail & I” coming out shortly. It’s like selling hippie wigs in Woolworth’s — National Post

“The average American has 79 sit-down meals in restaurants per year, 16% fewer than 15 years ago” — Fast Company

Supertax on bankers failed, says Darling — FT

“Is there not a senior gay in Apple design to throw up his hands and send a design like this back?” — Awl

Consumers don’t get the Times’s scoops. Advertisers are deserting it. Even celebs aren’t giving it interviews — Independent


Just yesterday morning I was thinking of the scene where they tried to stuff the chicken into tea pot.

Posted by grumblecakes | Report as abusive

A unified theory of New York biking

Felix Salmon
Sep 3, 2010 07:06 UTC

Most bikers in New York have their fair share of road rage. Commuting by bicycle in Manhattan has many things to be said for it, but it’s certainly not relaxing. And bicyclists as a group have surprisingly little public support. The question is, why? And I think I’ve worked out something approaching the Unified Theory of New York Biking:

Cyclists get no respect as road users. Instead, tragically, they’re treated like pedestrians.

Basically, people in New York — and I absolutely include the bicyclists themselves here, who are actually the worst offenders — start with the long-established interplay between pedestrians and motorists, and then layer on bicyclists as though the bike was a cool toy given to a suburban 13-year-old kid, rather than an efficient way of using the city’s streets as a way of getting from A to B.

Let’s take all the different permutations in order. To begin with, there’s the old bike-free status quo, where the possible interactions are pedestrian-pedestrian, pedestrian-motorist, or motorist-motorist. It’s worth thinking about these a bit, because they’re deeply ingrained in us, and they’re responsible for shaping the way we see everything else.

The pedestrian-pedestrian encounter is both chaotic and benign, just so long as you don’t work in the middle of Times Square. (Ahem.) People move slowly enough that they have lots of time to maneuver around each other as necessary, and most of the time, with the help of a little eye contact, large numbers of people are extremely good at walking with and around and across each other.

The motorist-motorist encounter, by contrast, is very highly choreographed, with lights and lanes and speed limits and indicator lights and even a dedicated corps of traffic police to enforce the rules. The rules aim to minimize car crashes, and again, as a general rule, they do a pretty good job.

Finally there’s the pedestrian-motorist encounter, which is based largely on asymmetry: motorists have nothing to fear from pedestrians, but pedestrians have everything to fear when it comes to getting hit by a car. At the same time, their respective spaces (sidewalk, roadway) are very clearly delineated, largely to minimize any need for the two to interact at all. When they do interact, pedestrians take advantage of the rules of the road: a red light, for instance, means that the cars have to stop, so pedestrians can cross against them. Pedestrians trust the motorists to follow the rules, and most of the time that’s what happens.

There are rules governing pedestrian behavior too, but they’re broadly ignored. Because they’re slow and harmless, pedestrians feel as though they have few responsibilities to others. So they’ll jaywalk, or cross in the middle of the block, or wait for the light to change while standing a couple of yards into the street, because they can. At the margin, a few motorists will be inconvenienced, but they have all the advantages of being in a car, so pedestrians feel it’s a fair trade-off.

The trouble all starts when you drop bicyclists into the mix. At that point, a whole new set of combinations comes into play, and as a city we haven’t worked out how to make them work. In other cities, especially in places like Copenhagen or Utrecht, bicycles are ubiquitous and everybody knows how to behave on and around them. But we’re not there yet.

Bikes can and should behave much more like cars than pedestrians. They should ride on the road, not the sidewalk. They should stop at lights, and pedestrians should be able to trust them to do so. They should use lights at night. And — of course, duh — they should ride in the right direction on one-way streets. None of this is a question of being polite; it’s the law. But in stark contrast to motorists, nearly all of whom follow nearly all the rules, most cyclists seem to treat the rules of the road as strictly optional. They’re still in the human-powered mindset of pedestrians, who feel pretty much completely unconstrained by rules.

The result is decidedly suboptimal for all concerned, but mostly for the bicyclists themselves. New York needs to make a collective quantum leap, from treating bicyclists like pedestrians to treating bicyclists like motorists. And unless and until it does, bike relations will continue to be marked by hostility and mistrust.

Consider the bicyclist-bicyclist encounter, first. Most of the time, bicyclists get on just fine with each other: we’re all riding along the street in the same direction, and if you need to do it, overtaking is pretty easy. You look behind to check for cars, you might announce a polite “on your left”, and off you go.

But all of that falls apart with the introduction of the evil bike salmon, which have reached pandemic proportions in New York, even on insanely busy avenues. If you’re riding the wrong way down the street, that’s always going to be dangerous for any bicyclists coming towards you. Sometimes, it’s downright lethal. I bike up Sixth Avenue to work, which nominally has a bike lane running up its left-hand side, but like all bike lanes this one is often filled with large opaque trucks. So I need to look behind me, merge into traffic, and skirt around the truck. All of which is no big deal, just so long as I don’t run headlong into a suicidal bike salmon coming the other way, who of course I couldn’t see in advance because the truck was in the way.

Or any other corner works much the same way: a friend of mine got some pretty nasty injuries when he turned a corner on a bike only to see a bike salmon of the delivery-boy subspecies barreling towards him. He slammed on his brakes, went over the handlebars, and the bike salmon went merrily on his way.

What justifies bike salmoning? Nothing. But what explains it is that bicyclists are in the pedestrian mindset: rules don’t apply to them. Yes, having a one-way system means you’ll sometimes have to go a couple of blocks out of your way, but cars do that automatically, and most of the time they’re going slower than the bikes. And none of this, of course, explains those delivery guys, who only bike the wrong way down the street. That’s just perverse.

Recently I saw a mother in her late 20s, riding down Avenue A with her toddler in a bike seat on the back. The mother wasn’t wearing a helmet, but she was wearing iPod headphones. And she was salmoning, which actually takes some doing on a two-way street like Avenue A: she was riding north, but on the west, southbound, side of the road. And she did this for a few blocks.

Now think of the message that mother was sending to any cars travelling south on Avenue A. It’s unambiguous: “I act like a pedestrian, I follow no rules, I don’t care about you, and you just have to navigate around me.”

Every bike salmon constitutes an utterly gratuitous confrontation and escalation in the war between bicyclists and motorists. Whenever a motorist encounters a bicyclist riding towards them on the street, that only serves to confirm in their mind that bicyclists aren’t proper road users, aren’t worthy of their respect, and certainly can’t be trusted to play by the same rules that govern cars. Bicyclists are an obstacle, an inconvenience — something which really shouldn’t be on the road at all.

As a result, drivers don’t treat cyclists as legitimate users of the road, even when they’re going in the right direction. Instead, they treat us as they would treat pedestrians. I’ve had a taxi driver scream at me for biking the right way down the street, because there wasn’t enough room for him to overtake and he wanted to get to the red light at the end of the block a few seconds faster. Once we were both stopped at the red light, he explained in a very forthright New York manner that he had every right to drive as fast as he wanted on the roads, and I had no right to be on the road at all.

I had much the same experience today — this is the one which prompted this entire blog entry. I was riding down 43rd Street to Reuters on the right-hand side of the street, passing cars waiting for the light to change at Broadway. Suddenly, a man threw open the rear door of one of those cars, right in my path; I slammed on the brakes and came to a halt, thankfully unharmed, just as he was getting out of his car. He didn’t apologize, so I smiled — I’ve learned that any sign of anger is counterproductive in these situations — and said that it’s always a good idea to look first, before opening a car door into the street.

His response was both nonsensical and illuminating: he informed me in a very haughty manner that I shouldn’t have been biking on the street in the first place. Confused, I looked around: did he mean that there was a bike lane I should have been using instead? No, there wasn’t. So I asked him what he meant, and he ignored me, rushing into 1500 Broadway.

What I think he meant, if he could articulate it — which clearly he couldn’t — was that bicyclists aren’t legitimate road users, and we shouldn’t be getting in the way of cars, or, for that matter, in the way of people exiting cars. No one worries about dooring pedestrians: for one thing, pedestrians don’t have the requisite velocity, and for another thing they’re not meant to be in the road in the first place. And bicyclists, in this guy’s mind, belong in the same category as pedestrians, not the same category as cars. (If there were enough room on the right for a car to pass by, you can be sure he’d look first before opening that door.)

You see that mindset all the time, with cars — especially when it comes to blinking. They’ll indicate for the benefit of other cars, but never for the benefit of bicyclists: if you’re switching into a new car lane, then you’ll blink, but if you’re going to turn across a bike lane, you won’t. All too often, they’ll commandeer bike lanes for themselves, turning them into de facto left-turn lanes. If it’s on the road, it’s for cars. And, of course, if they’re not using the bike lane to drive in, they’re using it to park in.

And while cars are reasonably polite, even in Manhattan, when it comes to cutting off other cars, they seem to have many fewer compunctions when it comes to bicyclists: they’re perfectly happy to zoom past me and then pull over to the curb right in front of me, forcing me to brake hard and try to maneuver around them. After all, they can do that with pedestrians, and no one minds.

Pedestrians can also navigate obstacles in the street, like those big metal plates or nasty potholes, a lot more easily than bicyclists can. We like very much to travel in a straight line when possible. But you should never assume, if you’re zooming along in a car, that the bicyclist you’re overtaking is going to remain in a perfectly straight line and that you can therefore overtake with only a few inches to spare. Any number of things can cause us to swerve unexpectedly — but drivers, at least in New York, often don’t remember that, or think that way.

If relations between motorists and bicyclists are bad, though, they’re nothing when it comes to relations between bicyclists and pedestrians. That relationship is positively poisonous, precisely because both sides are thinking of bikers as being more like pedestrians than like cars.

Why do bicyclists ride on the sidewalk? Because they think they’re pedestrians. And in doing so they infuriate the real pedestrians, who deserve the sidewalk to themselves. And while the majority of bicyclists don’t ride on the sidewalk, most of them do happily sit right in the middle of the pedestrian crosswalk. There’s no culture in New York of bicyclists giving way to pedestrians, and of stopping behind the crosswalk where they’re meant to stop. Instead, when they want to cross the street they do exactly what they do when they’re walking, and go as far as they possibly can without being run over by traffic. In doing so, they can get in the way of dozens of people just trying to walk across the street — and indeed even get directly in the way of fellow bicyclists coming up a bike lane towards them. Bicyclists always seem to forget how long their bikes are: they block off a lot of space, if you’re trying to cross past them.

Armed with their pedestrian mindset, bicyclists are convinced that they can cut easily through people crossing the street, just as they could if they were walking. They’re wrong, of course, but there’s no culture of giving way to pedestrians, because they feel even more defenseless than the pedestrians when it comes to the rough streets of New York City. And potential victims find it very hard to stop and think of themselves as being too aggressive.

Meanwhile, the obliviousness on the other side is utterly exasperating for any cyclist. I was riding down 44th Street recently and saw a guy wanting to cross the street mid-block. He looked at me, we made eye contact — and then he stepped out, right into my path! The point is, he was looking for cars, not for bikes. He saw me, but he didn’t think of me as a vehicle he shouldn’t step in front of; instead, he thought of me as a pedestrian who could get past him no problem.

While pedestrians are worried about cars running them over, and tend not to step out in front of them, they have no such compunctions when it comes to bikes, or bike lanes. Bike over the Manhattan bridge at any time, day or night, and you’ll find pedestrians walking happily on the north side, which is for bikes only, rather than on the much nicer pedestrian-only south side. I like to think that they simply have no idea of how much trouble they cause cyclists: the idea that they do know, and choose to walk in the bike lane regardless, is just too demoralizing to contemplate.

And the situation in some bike lanes — especially the one running down Broadway north and south of Times Square — is much, much worse, to the point at which the bike lane is actually unusable by bicycles. It’s painted green, and it’s set off from the street by a pedestrian zone, which means there’s no car danger at all, and which also means that pedestrians feel free to wander across it at will. And they never look first to see if a bike is coming. The bike lane essentially becomes an extension of the pedestrian zone, and the bikes are forced to use the road, defeating the whole point of building a bike lane in the first place.

One part is particularly bad: bikes are meant to be able to get down Broadway between 35th St and 33rd St, even though cars can’t. But no one seems to have told the pedestrians, who happily plonk chairs down in the middle of the narrow bike path between 34th and 33rd. It’s by far the shortest way for me to get home from work, but I always go well out of my way to take 9th Avenue instead: navigating the pedestrians on Broadway is just too hard.

Again, the problem here is mindset. The pedestrians are in a pedestrian mindset, where they can wander happily wherever they like, especially when there aren’t any cars to worry about. It simply never occurs to them that they might be getting in the way of bicyclists — even when they’re standing right in the middle of a bike lane. If cars use bike lanes as left-turn lanes, pedestrians use them as staging areas, places to stand while they’re waiting for the light to change.

Pedestrians intuitively understand that bike lanes are relatively safe from cars, and therefore feel safe stepping out into them without looking first. And that can be extremely dangerous, both for themselves and for cyclists: a friend of mine died after a pedestrian stepped out in front of him when he was riding his bike.

Bicyclists aren’t like pedestrians: we’re much faster, we can’t stop quickly, we can’t navigate as adroitly, and it takes a lot of effort to slow down and speed up again, compared to the effort expended in just moving at a constant velocity. We’re a danger to pedestrians, but they’re a danger to us, too. And cars, of course, are a danger to both of us.

As New York becomes an increasingly bike-friendly city, it’s going to have to how learn to deal with these new encounters: bike-bike, bike-car, bike-ped. Other cities have managed it; we can too. But for the time being, bicyclists are being thought of in the “pedestrian” bucket. And that’s causing a great deal of harm.

Update: Some great comments below. One thing is worth clarifying: I’m certainly not saying that bikes should behave exactly like cars, which would include not overtaking cars in their own lane. In fact, under New York State law it’s illegal for bikes to behave exactly like cars: if you’re on a road without a bike lane, you have to stay to the right of the road and let cars overtake you if possible. Overtaking  within a single lane of traffic isn’t just sensible, it’s the law!

Also, Caleb Crain found some pertinent statistics in Jeff Mapes’s book Pedaling Revolution:

According to Mapes, a 1996 study by the University of North Carolina Highway Safety Research Center found that “as many as a third of all bike accidents involved simply riding against the flow of traffic,” and a 2003–2004 Orlando, Florida, study found that “nearly two-thirds [of bike accidents] involved riding on the sidewalk or another unsafe choice by the cyclist.”

Update 2: After an interesting back-and-forth with Ledbury22 in the comments, he introduced himself to me as we were standing in line at the hardware store this afternoon. His problem is with “lane splitting”: when bikes create their own mini-lane between cars and the sidewalk. It’s built into New York law, which requires that bikes create just such a mini-lane on the right-hand side of the road even when there aren’t any slow or stationary cars, so that faster cars can create their own, much bigger, mini-lane, and overtake them. And it turns out that Ledbury dislikes this not from the perspective of a car driver, as I had assumed, but rather that of a pedestrian.

The problem, as it turns out, is a common one: pedestrian wants to cross a street where cars aren’t moving: they’re waiting at a light, or stuck in a jam. So walks out into the road in the middle of the block, without looking, and gets whacked by a bicyclist. Pedestrian’s fault, clearly — but when you’ve just been hit by a bicycle, you’re liable to start blaming bicycles for increasing the danger quotient on the roads even when you are the person at fault.

There’s really only one answer for this: pedestrians need to get used to the idea of looking for bikes just as they look for cars. This is one are where improved bicyclist behavior can’t help. Even increased bike lanes wouldn’t help much, since in my experience the kind of pedestrians who step out into the road without looking are even more likely to step out into a bike lane without looking.

Update 3: Bike Snob responds! And I agree with everything he says. His conclusion:

Salmon makes many excellent points, but I was dismayed to see he fell into the same trap (or, in his case, net) as most other people who try to address this issue, which is to suppose that drivers and cyclists and pedestrians are somehow “different,” or that their nature is somehow determined by their vehicle. Excluding for the moment the fact that many people are pedestrians and cyclists and drivers at various points in the day, a considerate person is a considerate person and an idiot is an idiot, and both will behave as such regardless of how they are propelling themselves at any given moment.


Great article! As someone who has done their fair share of being a cyclist, motorist, and pedestrian in NY, I think pedestrians could stand to hold themselves a lot more accountable. I personally have made it a point to always stay on the curb when waiting for lights (bc when you’re driving or biking, a narrow lane is a harrowing, frustrating lane), and I think that one act alone could diffuse a lot of road rage.

Also, how about some wrong way signs in bike lanes as a pre-emptive measure? I really think some people just aren’t aware and I think we should attempt to educate before we jump to penalize wrong way bikers. When I was in Seoul last year, every bike lane had wrong way painted in the lanes (facing the biking salmon), seemed very effective.

Posted by bangzy_ | Report as abusive

Microfinance datapoints of the day

Felix Salmon
Sep 2, 2010 21:06 UTC

Here’s three takes on microfinance, all from the past month:

  • Indian microfinance lender SKS goes public, raising $358 million and making its founder dynastically wealthy. The decision was controversial, and was largely responsible for an entire non-profit organization, Unitus, disappearing. When that kind of money is at stake, noble non-profit principles have a tendency to evaporate.
  • Steven Schwarcz of Duke University, has a bright idea: why not use the magic of securitization to provide funds for microfinance lenders? “Such disintermediation,” he writes, “would enable microfinance loans to be funded directly from low-cost, and virtually limitless, capital market sources”. What could possibly go wrong?
  • Hema Bansal reports from a roundtable in New Delhi, which found that many microfinance lenders don’t know their effective interest rate, and the ones who do know it tend not to reveal it.

The lesson I take from all of this is that microfinance is much, much messier than its advocates normally like to admit. I’m constantly astonished at the number of people I meet who are all excited about some microfinance fund or initiative, with some vague idea that the main problem is finding people to loan out dollars at low interest rates; after that, the magic of the market will help bring billions of people out of poverty.

But in reality it doesn’t work anything like that. Microfinance loans are often punitively expensive, with the main beneficiaries being the owners of the MFIs rather than the borrowers. MFIs which start off on a non-profit basis convert to being for-profits, following the well-trod path most famously trod by Compartamos, in Mexico. Insofar as default rates are low, that’s often because borrowers keep on rolling over their loans into ever-larger debts, without ever really exiting from poverty. The lenders compete viciously in densely-populated urban centers, being very opaque on pricing, and leaving many rural areas untouched. And all the while well-intentioned people in the US dream up new ways of throwing more dollars at the industry, in the absence of any evidence that more dollars is what the industry really needs, at this point.

My feeling is that good non-profit microfinance organizations do exist, and that they should be supported with grants first, with technical expertise on things like underwriting and growth strategies second, and with local-currency funding third. If someone tells you that you can help bring millions of people out of poverty while still making a profit on your investment, your first reaction should be that they’re selling something which is too good to be true. Start by asking about local interest rates and then work out from there: the facts on the ground are likely to be a lot more complicated, and a lot less compelling.

Update: Accion’s Andrew Sprung responds in the comments.


Yes, “the facts on the ground are likely to be a lot more complicated.” But Felix, it’s you who are oversimplifying, and overgeneralizing from a few overheated microfinance markets. It’s true that in some regions in India, and in other countries including Nicaragua — and years ago, Bolivia — too many MFIs piled into a limited market and too many clients took out multiple loans. But some facts that you don’t consider:

1) Rates that seem high by western standards on an annual basis are generally far lower than rates the poor in those markets are accustomed to paying to lenders from a variety of informal sources (those sources are not always exploitative. A one-week $10 loan with $10.25 due at week’s end comes to 261% on a compound annual basis (Portfolios of the Poor, http://amzn.to/bORHOC). In many cases, too, saving is as expensive as borrowing, as people pay substantial fees to have their money kept safely and delivered predictably.

2) Microfinance rates are driven mainly by the high cost of servicing lots of small loans. Nonprofits do not generally charge lower rates than for-profits.

3) MFIs cannot sustain themselves over time by inducing clients to roll over debt. If that were the norm for Compartamos, why does its client base keep expanding, and why are its repayment rates among the best in the business? No one is putting a gun to Mexicans’ heads and forcing them to take out loans.

4) The largest MFIs in India have moved quite swiftly to address problems of client overindebtedness that have cropped up in some torrid markets (while large swaths of India remain underserved). They have formed a trade association, MFIN, with 42 members, which is creating a credit bureau, with which it will require members to register, and has promulgated a comprehensive code of conduct that it plans to enforce strictly http://bit.ly/aD6mN6. A global effort, The Smart Campaign for client protection in microfinance, http://www.smartcampaign.org, / aims to institutionalize client protection principles throughout the global microfinance community

5. It is true that some maturing microfinance markets are not the most productive targets for aid dollars. Grant funding should be directed at MFIs that reach marginal groups, or for R&D for development of new products like microsavings and insurance, or for development of industry infrastructure like credit bureaus. See Congressional testimony by Elisabeth Rhyne, director of the Center for Financial Inclusion at ACCION http://bit.ly/d4scrO (disclosure: I work for ACCION).

5) For-profit institutions have not only helped lift people out of poverty- they are the primary engines of human prosperity. Investing in the right for-profit ventures can indeed help to alleviate poverty.

Posted by adsprung | Report as abusive

Why houses are like dishwashers

Felix Salmon
Sep 2, 2010 17:11 UTC

A lot of people have been quoting this passage from Chip Case’s op-ed on housing as an investment:

For people with a more realistic version of the American dream, buying a house now can make a lot of sense. Think of it as an investment. The return or yield on that investment comes in two forms. First, it provides what is called “net imputed rent from owner-occupied housing.” You live in the house and so it provides you with a real flow of valuable services. This part of the yield is counted as part of national income by the Commerce Department. It is the equivalent of about a 6 percent return on your investment after maintenance and repair, and it is constant over time in real terms. Consider it this way: when Enron went belly up, shareholders ended up with nothing, but when the housing market drops, homeowners still have a house. And this benefit is tax-free.

I read it a lot of times, and couldn’t really understand what it meant. So I decided to phone up Professor Case to ask him.

There are two things here which are easy to misunderstand. Firstly, when Case says “think of it as an investment”, he’s not using the word “investment” to mean “something which holds its value over time and which with any luck will actually appreciate in value”. Instead, at least in this passage, he’s thinking of an investment in much the same way as you might consider a dishwasher, or any other durable good, as an investment. You pay a sum of money up front, and then you get a stream of valuable services more or less in perpetuity. In the case of a dishwasher, that means clean dishes; in the case of a house, that means shelter in the place you want to live.

(Incidentally, that’s why I think Ryan Avent’s criticism of Case is misguided: the location is just as much a part of the service that you’re receiving as the shelter is.)

Secondly, when Case talks of this part of the investment being “constant over time in real terms”, he means something almost tautological: the value of any given service is constant in real terms by definition. What he doesn’t mean is that you’re going to get a constant 6% return on your investment after maintenance and repair. That number, he’s happy to admit, can rise and fall over time, and fluctuates mainly with the market rent for your home.

Essentially, what Case is saying here is that once you own a house, you’re guaranteed to be able to live there as long as you like, and that guarantee has real value. If what you want from any investment is an annuity — the ability to get constant real value out of it indefinitely — then in that sense a house is an investment. But that’s emphatically not the same as thinking of a house as an investment in a mark-to-market sense of how much you could sell it for.

Case is a long-term bull on housing: “sooner or later”, he says, prices will rise. Well, yes. But if it’s later rather than sooner, and if they fall a lot from current levels before they start rising, then you might not take much solace in the fact that you’re still getting the same kind of value out of your home that you’re getting out of any other durable good.


Of COURSE a house is an investment. What else would it be? You have cash flows out (down payment, principal, interest, taxes, upkeep) and cash flows in (imputed rent and sale price). You make your best guesses, plug in the numbers, and it’s a perfectly ordinary, Finance 101, time value of money computation. That is what an investment looks like. That is what an investment is.

Owning your own home is precisely equivalent to being your own landlord. Take a rent payment out of your right pocket and put it in your left pocket on the first of every month if it helps with the visualization.

The problem is that people have expected the return on housing to come from one part of the equation–resale value–and have expected that return to be very large and absolutely certain, when it is neither. The return comes mostly from the imputed rent: putting that money into your right pocket instead of someone else’s.

Now, housing may not be a GOOD investment–but that’s a different kettle of fish entirely. The S&P500 wasn’t a good investment over the last decade, but only a fool would say that it was therefore not an investment at all.

Posted by ckbryant | Report as abusive

When bankers are more dangerous than warlords

Felix Salmon
Sep 2, 2010 14:56 UTC

Amy Davidson has a good round-up of the tragic bank run at Kabul Bank, which is threatening not only the largest and most important bank in Afghanistan but also what remains of that country’s shattered political economy. But you can pretty much learn everything you need to know from reading two sentences from the excellent WaPo report:

Speaking Wednesday from his villa in Dubai, which was paid for by Kabul Bank, Mahmoud Karzai, the president’s brother, said cash withdrawals from the bank were a “little bit more than usual” but did not threaten to cause a meltdown. A full-scale run on Kabul Bank, he added, “would be a major disaster.”

Yes, the president’s brother is a part owner of the bank, and he’s living in Dubai, in a villa paid for by the bank — which, incidentally, handles the payroll for Afghan soldiers and schoolteachers — and really, what could possibly go wrong?

As the FCIC revisits the Lehman Brothers failure for the umpteenth time, the U.S. is faced with yet another decision about whether or not to bail out a bank whose failure could have enormous systemic consequences. Is there any precedent for one country bailing out another country’s bank? Would the money come from Treasury or from the Pentagon? And how on earth would such an action play in the midterm elections?

It’s all extremely fraught, but the conclusion to the WaPo story pretty much sums it up, I think.

One senior Afghan official, who spoke on the condition of anonymity, said that he had hoped for the best but that “the worst is happening.”

America is certainly finding the military strategy in Afghanistan hard going. But could it be that the country will finally be undone by it bankers?


I think that the government will believe that it would be a good idea, but won’t do it, simply because of political suicide. What politician would bail out yet another bank that doesn’t take part in the US economy?

The main problem is that even if we were to bail them out the Afghans probably wouldn’t be that grateful anyway. Taxpayers would be furious, thinking that their money was ‘wasted’ on something that had seemingly no effect on the effort and their trust.

Yet if we don’t we will inevitably get blamed for its loss, like TaxLawyer said. I’m guessing a loss in Afghanistan might be inevitable if more lose-lose situations like this keep cropping up.

Posted by CPie | Report as abusive

Quants merge with humans

Felix Salmon
Sep 2, 2010 13:55 UTC

Eleanor Laise has found an interesting trend in the world of quant funds: a lot of them are looking much more human, these days.

Quants are seeking to win back investors not just with their characteristic number crunching but also with a bit of soul searching. Many of the funds’ managers are seeking to make their models a little more like people, by making them more responsive to changing circumstances. That can mean revisiting computer models more often, tweaking their components, or incorporating measures of macroeconomic risk rather than just stock-specific information.

Quant managers need to understand “that financial markets are better understood through the lenses of a biologist rather than a physicist,” says Andrew Lo, a finance professor at the Massachusetts Institute of Technology who also manages quant funds.

My feeling is that what we’re seeing here is the beginning of a kind of Hegelian synthesis in the fund-management world. The thesis was that a talented active manager, with experience and insight, could outperform the market. The antithesis was that active managers, being human, had a distressing tendency to do exactly the wrong thing at exactly the wrong time. Instead, computers had in reality the discipline that human traders have only in theory, and can stick to any given strategy through thick and thin, just as the backtesters intended.

The synthesis, in this view, is that it’s not always smart to stick to a failing strategy, although humans can definitely benefit from the sheer computational power embedded in quant models. So the humans use those models, to a greater or lesser extent, to inform their investment decisions.

At that point, it all comes together: quant funds are the ones which minimize human meddling, while other fund managers who would never consider themselves quants still use sophisticated models to help them pick stocks and strategies. But in reality they’re not so far apart.

I do think that Lo is right, and that it’s never particularly smart to simply stick to a single strategy in an attempt to outperform the market. (In fact, I’m not a big fan of even trying to outperform the market in the first place, but that’s a separate question.) On the other hand, given that human tendency to do the wrong thing at the wrong time, I fear that human-inflected quant funds are just going to end up abandoning strategies just when they would have started to work.

What I’m not seeing, anywhere, is a dynamic quant strategy which automatically makes significant changes to its investment style depending on market conditions. Quant funds tend to be perfected by humans and then released into the wild; any further changes, at that point, are also performed by humans, who don’t trust the computer model to optimize itself on the fly. Sure, computers can buy and sell stocks as conditions change. But they don’t change the rules governing which stocks to buy and sell: those are fixed unless and until humans change them.

That’s probably just as well: I’m not sure the world really needs investment strategies being set without any fund manager having a clue what they actually are. But at the same time, a purely computer-generated strategy might provide some interesting diversification from the madness of crowds. I doubt we’ll see anybody admit to using one any time soon. But for all I know it’s already happening at some hedge fund somewhere. Maybe it’s even the secret of RenTech’s famous and mysterious black box.


Garbage in-garbage out. No more complex than that. Given an erroneous quant model, one will get erroneous results. Unless we develop AI that can change on the fly and learn from itself, we are just trading computer modeling inputs devised by humans, with human judgment. One is inflexible, and the other behavior-dependent human-instinct often irrational. At least the human manager can change strategy–the quants can lose unimaginable amounts of money chasing the inputs of its programmers.

Posted by TaxLawyer | Report as abusive