News Flash: Loans Are Risky
December 18, 2011 23 Comments
Although few know in any detail what the ‘Volcker Rule’ is, everyone who even aspires to hold anything resembling a semi-respectable opinion on The Economy is fully on board with the philosophy behind it, which is that banks shouldn’t engage in ‘prop trading’ or, more generally, ‘risky’ stuff. Here is a typical example, which I cite only as a useful instance of the eminent reasonable-soundingness of the position:
…legislative intent in this legislation is clear…to cut banks down to size and return them to their historical role of intermediating between savers and borrowers. [...] the Volcker Rule directs the regulators to get banks out of the business of betting on the markets.
Hear, hear! you all say. It’s about time! Banks shouldn’t bet on markets. The horror! They should just do the pure and safe and vanilla ‘intermediating between savers and borrowers’ that our grandpas’ banks did.
Being (as far as I can tell) the only person on the entire Internet who doesn’t think ‘prop trading’ should be banned, and therefore is against the Volcker Rule or anything like it lock stock & barrel, allow me to briefly explain what is wrong with the preceding conventional wisdom. Because it is, in fact, a contradiction.
How exactly do banks ‘intermediate’ between savers and borrowers? A saver is someone who wants to make a 1-day loan to a bank. A borrower is someone who wants to borrow 5-, 10- or 30-year money from a bank. How exactly can you ‘intermediate’ between two such people?
The short answer is, ‘you can’t’. The saver and the borrower are not coming to you with truly offsetting positions. So the real answer is, by taking risks.
The above, pure, clean, grandpa-approved banking transaction entails huge risk! What on earth could be more risky than borrowing 1-day money (from the savers) and turning around and lending it out long-term? You’ve all seen It’s A Wonderful Life; what if the savers all come to your teller window at the same time and want their money back? Separately, what if the borrowers’ loans come due and turns out they’re bankrupt? What if the assets the borrowers pledged as collateral (bonds, houses, whatever) suddenly drop in value? What if interest-rates precipitously rise and you suddenly find yourself having to pay savers far more to borrow from them than the average rate you’ve already locked in making all those long-term loans?
If you want banks taking deposits and making loans, then by definition you DO SO want them taking risk. If you don’t want them taking risk, you don’t want banking at all. Modern banking, or no-risk: pick one. The people who complain that banks ‘took risks’ and don’t want them ‘betting on the markets’, yet in the same breath say they should ‘merely’ be focusing on making fractional-reserve maturity-mismatched loans with depositors’ money (!), are either just ignorant or haven’t thought through the situation enough. News flash, all you Conventionally-Wise: LOANS ARE RISKY.
What? Oh, I see, the risk in vanilla lending is ‘understood’ and can be regulated and ‘controlled’ and, like, hedged because we have so much experience with all that sort of thing. Right, I forgot. I guess that’s why all those mortgages (a very, very vanilla and traditional and grandpa-approved type of loan) in the past 10-15 years have all worked out so well and caused no problems for anyone.
Huh? Oh, the mortgages only caused problems because of evil derivatives like mortgage-backed securities and CDS? Um, you do realize that that’s how banks hedged those mortgages, right? If you wanted banks controlling/hedging their risk – and you all sure talk like you do – then CDS was precisely what you were asking for. Similarly, banks can control the risk in their loan portfolios by hedging them – with derivatives. They can hedge the interest-rate risk – with interest-rate swaps (more derivatives). Those derivatives lead to more risk (i.e. counterparty risk), which can be hedged – with more derivatives.
So I could (theoretically) concede this conventional-wisdom notion that grandpa-banking risks can be wisely controlled and hedged, but at the cost of you acknowledging that all sorts of derivatives will need to exist because they play such a crucial role in hedging and spreading the risk inherent in the banking sector. But once we admit that we’ve stepped well away from savers-and-borrowers grandpa-banking and have headed pretty far downfield towards (gasp) ‘prop trading’. After all, the question is no longer whether banks will take risk (they will – loans are risky!), or stick to the ‘good’ sort of grandpa-approved instruments like loans (they won’t – because they can’t – because if they did they wouldn’t be able to control the risks you say you want controlled). The question is simply which risks banks will take, and how much. One way or another, you’ve already conceded that banks will have to be allowed to have desks trading, like, CDS and the CDX.NA.IG credit index and 10-year interest-rate-swaps and suchlike. That desk, whatever it does, will always have taken on some amount of risk or basis. What level of said risk you choose to call ‘prop trading’ is a matter of boring semantics. To borrow from the joke about the high-class lady who protest she’s not a prostitute, we’ve already established that banks will take on risk and therefore (intentionally or not) make bets on the market – we’re just haggling about the amount… That is because at the end of the day there’s no actual ‘bright line’ you can actually draw between ‘good’ risk, and evil ‘prop-trading’ risk. You might think there is, and think that regulators are smart enough to draw it, but you are wrong.
Now, I am just as against private gains/socialized losses as you are. But this ongoing project to micromanage, at the regulatory level, which and how much risk banks can take is an absurd and doomed one. Instead, I am here to say, on the Internet, that we should let banks ‘prop trade’ all they want. BUT, we should also let them fail, if/when they become insolvent doing so.
In fact, if we did that, and credibly, we might find that banks would do a far better job policing their own risk than would regulators sitting in Washington.
“you’ve already conceded that banks will have to be allowed to have desks trading, like, CDS and the CDX.NA.IG credit index and 10-year interest-rate-swaps and suchlike. ”
Why?
If you don’t let them hedge their risk on vanilla mortgages using derivatives, they will simply write fewer vanilla mortgages to control their risk. Banks in fact issued mortgages long before we invented CDS or interest rate swaps. A lower level of mortgage issuance would be a more efficient market.
If you don’t let them hedge their risk on ‘vanilla’ mortgages (which are, at least in the U.S., rather complex embedded options) using derivatives, you aren’t letting them hedge their risk. Hence, they will be (gasp) taking risk!
Maybe that’s fine but it flies in the face of the ‘I don’t want banks taking risk’ rhetoric used to justify the Volcker Rule. Do Volcker Rule supporters want banks taking risk or don’t they? They, not me, need to make up their minds.
You are just being a little nit.
People want banks to take clearly understood risks, based on simple financial products, in quantities that make it unlikely that adverse variance will eat through thier equity. The simpler the products, the fewer actors involved, the more likely variance can be understood and controlled such that banks don’t go boom.
Vanilla mortgages, without all the derivatives shinanigans, are pretty easy to understand variance wise. We’ve been doing that a long long time. Without all the bells and whistles its pretty easy to design a bank where adverse variance on vanilla mortgages, issued and held by the bank in question, can be brought in line with the equity cushion of the bank.
What they don’t want banks doing is taking complicated massively leveraged bets on arcane financial products and then getting their balance sheets all gunked up with every other bank.
“But writing fewer mortgages is the opposite of what Volcker Rule supporters say they want”
Well, its not the opposite of what I say I want. I want fewer mortgages written.
To your credit, your position is consistent. You favor a banking system that is limited in certain ways, and you recognize that this would inevitably lead to mortgages that are more difficult to get and more costly for the homebuyers. This (and similarly, fewer small-business loans) – thus, fewer jobs and fewer homeowners – is indeed the sort of outcome one must expect from an intellectually-consistent and well-enforced ‘Volcker Rule’.
I wonder, however, how many people who say they favor the ‘Volcker Rule’ realize that this is what they are asking for? Again, the reason I say this is that many of the same people speak as if they want banks to lend more and that somehow the ‘prop trading’ they imagine to be the problem is getting in the way of that, because it’s ‘risky’. I’m glad to see that you and I agree that lending is risky too, it’s just that you want to reduce lending overall. Fair enough.
As to the details of the restrictions you favor, I still think we have a disconnect. You claim that people ‘want’ banks to use only ‘simple financial products’. I think you are completely wrong. The U.S. mortgage is anything but a ‘simple financial product’. It contains an embedded prepay option which gives the mortgage-writer a highly negatively-convex exposure to interest rates. To some extent it also implicitly contains a ‘default option’, which effectively means the lender has sold the borrower a put option on housing prices. These are among the most complex risks you can find in finance and managing these risks is anything but a ‘simple’ proposition. Yet, ask a cross section of the American public whether they ‘want’ to give up their prepay option (i.e. the right to refinance a mortgage when rates go down), and I’m pretty sure the answer is a resounding no. Meanwhile, there’s the fact that the entire market is completely distorted since the ’30s by the mammoth presence of the GSEs.
And so if you mean to include such mortgages among the ‘clearly understood’ risks, well perhaps you understand those risks better than I do. That is quite possible since I still grapple with them. I would dispute however that (aside from yourself) they are as ‘well understood’ as you make them sound; plenty of banks have lost money on these and related risks in mortgages as you probably know. ‘Unlikely’ variances are a funny thing – they happen more than everyone thinks. Before they happened, anyone would have considered the interest-rate environment of the late ’70s to be ‘unlikely’; anyone would have said that for home prices to drop 20-30% from the mid-2000s peak would be ‘unlikely’; nevertheless they both happened and those things – FAR MORE THAN ‘prop trading’ – caused major problems. So I am skeptical of relying on ‘unlikely’ as your risk management technique.
Ultimately you are saying you would control all these risks by just having banks do less of them – less loans, less mortgages, less everything. That is certainly one way. But it’s not clear that it’s the best or most efficient way (and it’s certainly not what most people think they are asking for with a Volcker Rule). What is? I don’t know! I would leave it to the individual banks to sort that out. The only counterargument to doing so is that the taxpayer might end up picking up the cost of saving banks. My answer: “let’s not, then”.
Yes, the 30 year mortgage is a government creation that probably shouldn’t exist. I would not mourn its passing.
Yes, the governments desire to get something for nothing led to price destabilization that largely made the regulated model for banking obsolete. In the 60s when we had to make hard choices as a society we did what democracies do best, not make them. So we ended up with the 70s inflation. Then when we wanted no more of that but still didn’t want to make hard choices we took actions to give us the “great moderation”, better known as shifting all of that wage inflation to asset inflation via debt. Its pretty hard to run a 1950s style regulated banking model in a world with rapidly occilating prices thanks to Fed hyperactivity and government free lunchism.
So I’ll agree that to go back to olden times regulation of banks we need responsible monetary policy as well, and likely responsible government spending/debt policy.
I’m of the opinion that increased debt levels don’t help anyone. They don’t create additional real resources. Some debt is necessary so that new businesses can form and create new real resources, but that debt must represent current real resources that are going to go into that process.
Bidding eachother up in a game of debt chicken doesn’t help anyone. Is the world any better off because family A and family B engaged in ever higher debt/income levels so they could outbid the other family for the house in the good school district. At the end of the day, bidding up houses doesn’t actually increase the number of good school districts. A world in which interest bearing debt made up a smaller portion of the money stock would not be a bad one.
While we have plenty of common ground in our skepticism of debt and of government’s distorting role in the mortgage market, this has taken us pretty far afield of the topic.
Let’s take it as read that you and I successfully sweep our hands and (somehow) get rid of the 30-year mortgage and the huge role of debt in our economy. Now then, back to the topic at hand:
Do you want banks to exist? If yes,
What do you want them to do? If something like ‘mortgages and loans’,
Do you recognize that such activity is inherently risky and there’s no getting around that?
Because that is all I’m saying. There is an undeniable tension between a philosophy that screams ‘banks shouldn’t take risks!’ and yet envisions them doing, well, banking. The issue is not whether, it is simply how much.
The ‘Volcker Rule’ approach envisions that the ‘how much’ can be sagely controlled by financial-whiz technocrats who work out, Goldilocks-style, just the right amount of risk to allow banks to take, and impose it using the omniscience and omnipotence that we all know regulators have. I am here to suggest that this approach is retarded.
It’s all well and good to not want banks to take ‘too much’ risk (whatever that is imagined to mean), but I care much less if we don’t bail them out when they do. So, let’s not. I fail to see what objection there can be to this suggestion of mine.
best
“financial-whiz technocrats who work out, Goldilocks-style, just the right amount of risk to allow banks to take,”
And yet, people on the gaming commission can manage to do it for casinos. And insurance agenices do it reasonbly well with insurance companies (they mainly get in trouble when they try to act like investment banks and “innovate”).
In truth, regulators don’t have to be financial wizards if the stuff they are regulating is simple and has 30 years of historical experience. They mainly get in trouble when they can’t keep up with banks “innovation”. So the simplest solution is to cut down on financial “innovation”.
But its “innovation” you cry. No, it isn’t. Nearly all “creative” work in the financial sector is finding new ways to exploit principal agent problems (you know, those things that make markets inefficient). I consider 90%+ of what the financial sector comes up with useless. If you simply change your mindset from worrying regulators will stifle “innovation” to thinking that financial innovation is fundamentally less valuable then actual innovation (you know, making a real product or service), then it becomes easy. Companies can’t run loops aroun regulators if they are severly limited in their range of products and all new products start out guilty and must prove innocence. In the real economy sector such regulation might be stifling but in the financial economy sector its great.
“but I care much less if we don’t bail them out when they do. So, let’s not. I fail to see what objection there can be to this suggestion of mine.”
No politician will ever vote for bank liquidation. And any that did would be quickly removed by “the people”. You live in a democracy, you have to live with its political economy constraints. This goes deeper then the latest crisis and is pretty much true of all banking crisises.
There were much better ways to handle the bailout that would have been fairer and had better economic implications. But liquidation was never on the table because of political constraints.
If you arguement doesn’t have a solid political economy behind it then its a non starter.
I don’t know what to make of two bizarrely inapt analogies of financial markets to controlled casinos in one thread. I certainly hope you are same person?
In any event, you appear to have assumed I am against Volcker-Rule-ism primarily because I want/crave “innovation” in the financial sector. I don’t know where you got this idea (certainly not from something I wrote). I yield to no one in my ambivalence towards “innovation”, most of which (in my view) only comes about in reaction to (mostly dumb) government/regulator rules anyway. I would happily toss both the spiderweb of regulation, and the symbiotic “innovation” it inevitably engenders, overboard. (I wrote an old post about this. Basically, ask yourself, why do “CDOs” exist?)
No, the reason I am against Volcker-Rule-ism is far more basic. On some level I am just against idiocy in law and regulation. So there’s that. The mindset is also centralized and technocratic – Smart People can manage things from above – and the misguided hubris of that failed notion I view as being so well established as to be beyond dispute. So generally, because of its dumbness, cost, and lack of tangible effects, I see such regulation as parasitical on the economy – not the Goldman Sachs economy (about which I care little), but the real economy. I have seen plenty of evidence that such a regulatory approach leads to a zillion well-paid white-collar jobs for lawyers, accountants, financial whizzes, and other degreed upper-middle-class folk. I have seen precious little evidence that such regulation accomplishes anything tangible on behalf of the average citizen or taxpayer.
If you want a shorter version of all this, I would point you to a handy oft-repeated quote from Arnold Kling of Econlog: All these rules are essentially aimed toward making the financial sector ‘harder to break’. I would, instead, aim towards making it easier to fix when it does.
best
I can sympathize with the Hayekian sentiment and at the same time its utterly idiotic. We did manage to do this at one time. There was a window after the depression and before the 70s inflation when the regulated banking model worked really well. And its worked reasonably well for other sectors with stronger regulation (like insurance). Its not like its some pie in the sky theory that has never worked in the real world.
Your theory is that we can’t regulate, because regulation broke down over the last 30 years. But as best I can tell regulation broke down because we stopped trying to regulate well (in fairness, the inflation made this hard). We went from the post depression model where finance shenanigans is guilty until proven innocent, to the post Reagen model, where financial innovation is innocent until regulators prove it guilty. In the latter environment, regulators are not going to be able to do their jobs. That’s obvious. You can’t expect regulators to keep up with a private sector that outnumbers and outguns them by several orders of magnitude.
CDOs are a good example of this. In the innocent until guilty model they can be used to skirt regulations until they blow up. In the guilty until innocent model they never get introduced, because its obvious they are just about skirting regulations.
Now, guilty until proven innocent can be pretty stifling, but the financial economy isn’t the real economy. Stifle it all you want.
I have to say I’m a bit skeptical of the ‘everything was fine our grandfather’s banking’ story. Even if it’s true that there were fewer bank problems, the ‘window’ you speak of was characterized by many things (winning a world war, government distortion of the economy, baby boom) that don’t quite transfer as a recipe for repetition. One might just as easily say that Social Security ‘worked just fine’ at a 1% payroll tax level, so we can still go back to that. In any event you’re the one who (like me) wants to get rid of 30-year mortgages and (presumably) much of what Fannie does, so right there you have already tossed the ’30s-’60s grandfather’s-banking scenario overboard.
I never said we ‘can’t regulate’. I am all for regulation, I guess (I’m actually rather ambivalent towards it…meh), but in any event all regulation is not created equal. What I object to is not the existence of regulation per se but the way we are trying to regulate, which is characterized by total top-down control, an assumption of (i.e. demand for) omniscience about the inner workings of large organizations and their balance sheets, and an overestimation of the regulators’ abilities to manage/control risk and activities. I object to this partly because it’s stupid and pointless and wasteful (in the sense, again, of transferring wealth away from workers and to the white-collar yuppies who end up supposedly-enforcing it all). But I also object to it because it simply is never going to work. Not only is it never going to work, its potential for working, as society gets bigger and more complex, will inevitably get worse and worse (and you essentially say as much in your comment!). But as you know this is just a rehash of the problem faced by all central planning – which our current regulatory approach amounts to.
As for innocent-until-proven-guilty, well this is a hallmark of American jurisprudence and a prerequisite for the rule of law. The rule of law precludes that laws/regulations be vague. Financial people may be evil and greedy and you may dislike them, but they are still people. When you imply that they should be guilty-until-proven-innocent, the effects of what you are saying are that certain people should be thrown in jail for running afoul of laws whose boundaries were unclear to start with. I’m not sure you realize the import of what you are saying. It’s not just ‘stifling’ of ‘financial innovation’ you are describing (if it were I wouldn’t care!), it’s throwing actual people in actual jail and taking away their property for violating some ‘regulation’ that you purposefully want to keep vague.
A situation in which people engaging in mundane tasks can’t ever be sure the government won’t come along and decide to jail them for some random bullcrap (because of a regulatory approach that rejects innocent-until-prove-guilty) is definitely part and parcel of what I object to in the current regulatory mindset. Because frankly, I think we’re already there. I can read your comment as saying that you think that’s fine and dandy, or give you a chance to revise/clarify. Either way, I don’t think it’s fine and dandy.
And as for ‘just write fewer mortgages’, well yes, that is another way of controlling their risk. And maybe that’s fine. But writing fewer mortgages is the opposite of what Volcker Rule supporters say they want – after all, they say banks are supposed to ‘get back to’ the basic grandpa-banking of loans and mortgages. Indeed, Volcker cheerleaders are often the same people constantly lamenting that banks ‘aren’t lending’. They certainly don’t openly advertise ‘fewer and harder-to-get mortgages for everyone!’ as a benefit of the Volcker Rule approach….
What your harping about is meaningless. Casino’s take millions of risks a day, but they are simple well understood risks that they can control. And regulators make sure they have enough cash on hand to handle a bad day. Thus casinos don’t go boom much. To say they are taking risks misses the forest for the trees. Variance isn’t risky. Variance you don’t understand and can’t control is risky.
The risks in casinos are simple and understood because they come from highly contrived and controlled games whose statistics are, by construction and design, simple and predefined.
What this has to do with financial markets is beyond me. But it is precisely the ignorant and hubristic notion that these or those financial risks can ever be ‘simple’ and ‘well understood’ such that ‘regulators’ can ‘make sure they have enough’ that I mean to disparage here.
Do you really believe that casino games are that much simpler then writing a million mortgages to people with documented income and high down payments? Insurance products are even more complicated (mathematically) and yet insurance compaies manage to sell insurance without blowing up the way banks do. The variance is controllable, I’d say you are drawing up a useless distinction.
Now, if we are talking about loans to start ups, that might be harder, but we aren’t. We managed to fuck up the simplest fucking sector of the economy, housing, which should be easy as hell to control.
“Do you really believe that casino games are that much simpler then writing a million mortgages to people with documented income and high down payments?”
Let’s go all Socratic here.
What’s the covariance between spins of a roulette wheel?
What’s the covariance between the blackjack table and roulette?
Simply by running short, repeated, independent (covar = 0) games with thousands of trials per night the central limit theorem guarantees a normal distribution of returns. Guarantees – as in is a mathematical law. This isn’t simple math but it’s all well understood. The normal distribution perfectly describes a casino’s risks.
(The correct answers are both 0.)
Ok, on to housing and loans:
1) What are your financial risks if you make someone a collateralized loan on an asset that will either be paid off or satisfied by the delivery of the asset?
2) What is the covariance between the price of a house and the price of another house in the same neighborhood?
3) What is the covariance between the price of a house in New York and the price of a house in California?
4) What is the covariance between changes to the price of a house and changes to the income of the owner of that house?
5) Covar between housing prices and interest rates?
6) Covar between interest rates and borrower income?
0 is the correct answer to none of these questions and there are many many more.
Banking isn’t a simple model like casino games.
You can talk about the variance of casino games because you know the distribution (normal). You don’t even know the distribution of returns for banks. You can model and approximate but the “financial crisis” is a great example of how everyone making the same assumptions (making the same assumptions is, of course, legally required) leads to disastrous outcomes when the models diverge from reality.
This isn’t even close to the last word about banking and the financial crisis but you absolutely have to understand this to even begin to speak intelligently about it.
Do you really believe that casino games are that much simpler then writing a million mortgages to people with documented income and high down payments?
Um, yes. The former are basically physical implementations of predefined probability distributions that are known in advance and do not vary. The latter…aren’t even close. It is just not even close at all.
You’ve written a million mortgages at a 5.5% WAC. Now interest rates come down 1%. How many people are going to refi? How many of those refis are due to the natural background housing turnover? What if home prices as measured by the Case-Shiller index appreciate at a 2% annual rate? 5%? Do your answers change? Say these are non-agency mortgages, now say unemployment goes up 1%, how many are going to roll from 60-days-delinquent into the 90+ delinquent bucket? 5% of your pool is 90+, how many will become REO, what’s the average time it will take to liquidate them, and what is destined to be the loss severity? Do any of your answers change if I tell you 80% of the mortgages come from CA/NV/FLA/NV? don’t come from CA/NV/FLA/NV? were originated by Countrywide? weren’t originated by Countrywide? are serviced by this or that loan servicer? If your answers do change, then how exactly? If they don’t change, why not?
Are these really things you think are known in advance, the same way as is the probability of rolling 5 in craps and winning on that odds bet behind the Pass line? Are these things that you think make no difference whatsoever to the value of mortgages to the one who wrote them? Are these things that you think belong to static/invariant distributions that have no meaningful chance of blowing up on you in some direction (and thus can be easily regulated/managed simply by, like, appropriate 99.9%-confidence-level-based provisioning)? Any ‘yes’ answer here marks you as not knowing whereof you speak, and would also be quite puzzling because it would mean you don’t even recall what happened to the housing market as recently as, like, 3-4 years ago… Meanwhile ‘no’ answers means you’re conceding my point. So you tell me.
The other bizarre thing is that you’re so super-confident that these risks are so well ‘understood’ and can be so easily ‘controlled’ while at the same time you don’t want banks to use, like, interest-rate swaps (because they are [gasp] ‘innovative’ derivatives). Clearly caps/floors are out. And credit risk? Will you give your blessing to, like, shorting the ABX or CMBX indices against mortgages? Surely not. How do you even reckon they will manage these risks at all? Oh right – just don’t make loans in the first place.
That is certainly one way.
P.S. You do realize that insurance companies fail/are taken over by the regulator/are restructured all the time, do you not? Not sure why you keep saying insurance companies never blow up. Check recent news. What sort of company is The PMI Group, Inc. in your view?
Or, what Steve Johnson said
Pingback: Goldilocks Thinking « Rhymes With Cars & Girls
Pingback: The Professor Who Fixed Banking With Her Mind « Rhymes With Cars & Girls
Pingback: Ritholtz Doesn’t Understand Your Objection To Banning Something That Has No Definition « Rhymes With Cars & Girls
Pingback: Cracks In The Standard Model (Of Trading) « Rhymes With Cars & Girls
Pingback: Excellent Anti-Volcker Rule Editorial That I Couldn’t Have Written Better Myself (Because I Didn’t) « Rhymes With Cars & Girls
Pingback: I’d Rather Ask Paris « Rhymes With Cars & Girls