(Cross posted from
Moon of Alabama, the site where many of Billmon's Whiskey Bar commenters have found a new home.)
This is a bit long, and going sideways at times, but the final conclusion is very relevant at a time like today when Bush is proposing to hand over trillions of dollars to Wall Street to manage. So here goes:
1. There is a mathematically certain way to win when you play heads or tails (which can be applied to any similar game with results with known probabilities):
- you bet 1 on heads
- if you win, that's it, you can go back to step one
- if you lose, you double your bet: you bet 2 on heads
- if you win, you won 1 overall (lost 1, won 2) and that's it
- if you lose, you double you bet again: you bet 4 on heads
- if you win, you won 1 overall (lost 1, lost 2, won 4) and that's it
- if you lose, you keep on doubling your bet until you win - and you will have won 1.
Of course, the amount of time to win 1 is unknown. The other problem is that you must win before you run out of money. 1, 2, 4 sound like smallish amounts, but if you get to 32 and lose, you will need 64 to keep on playing, and then 128, etc...). So this is a certain way to win - provided that you have sufficient liquidity available.
2. Have you ever wondered why banks always have massive, plush, some times extravagant headquarters? It's not that bankers particularly need the luxury (although they do enjoy it), it's that they need to project an impression of wealth, because their survival depends on it. Banks manage and CREATE money and money is purely a trust business. Money nowadays is purely a convention - a dollar bill has value not because of any intrinsic value, but because everybody agrees that it has value and accepts it in exchange for real goods or services. That convenient role can disappear if people start doubting its value either as a instrument of trade or a a way to store value. They can doubt it for rational reasons (too much paper printed and thus too much money chasing too few goods - runaway inflation) or for irrational reasons (if it becomes "common wisdom" that money is not worth what it says it is).
Banks, as the institutions that create the money (by providing credit - when they provide a loan to you, you have "value" - money that you can spend, and the banks still has the same value for itself in its books - your obligation to repay is an asset which has value) are at the heart of what makes money "valuable". Banks' basic assets are the deposits from people - their first role is to store your wealth safely. The next role is to then use these deposits to lend some of that value to others while the first clients do not need it. This has evolved, with banks providing (a lot) more loans than they actually have in hard cash, because they know that their clients do not need to take their money out of the bank very often and they do not do it all at the same time. As long as you have enough funds to take care of normal withdrawal requirements, you can recycle the rest in loans. This provides a valuable service to the economy, by converting short term deposits (money that can be withdrawn freely at any time) into long term funds (loans repayable over a more or less long period). This allows investment and thus growth.
The flip side is that it is an inherently dangerous business - if every depositor wants his/her cash back, then the bank will not have enough money at hand in the short term to pay everybody back and may be forced to fold. That's a bank run, and it can be caused by a simple loss of trust without any objective cause, in a self-fulfilling prophecy (the simple fact that people believe the bank to be weak make them take their money out, which does weaken the bank). Trust is therefore essential for a bank to function properly - trust that it will be able to service normal withdrawal requests in full at all times.
The problem is that a run on a bank does not only damage the banks and its clients, but can have an impact on the whole economy. Loss of trust in a bank can always turn into a loss of trust in the banking system; insolvency or simply illiquidity (when it does not have enough cash at hand even though it has valuable, but not liquid, assets in its books) of a bank threatens all entities that have assets with that bank - depositors, corporate treasuries, other banks, etc...and that illiquidity can spread to otherwise healthy entities. The existence of such macro risks, called "systemic risk" has led governments to heavily intervene in the banking sector, either directly (by owning part or all of the banking sector) or by regulating it heavily. The main tools - deposits guarantees, minimum capital requirements, risk diversification obligations, etc... have been designed after big banking crises. Here, "big" really mens "big" - banking crises can typically cost 10-50% of the yearly GDP of a country to clean up.
The next problem arising form governments intervening to avoid systemic risk is that it encourages banks to take more risks, in the knowledge that they will be rescued if things go wrong. Heads, you win, tails, the government loses, why shouldn't you play? This is called moral hazard: it encourages reckless behavior precisely by making the consequences of such behavior less bad. This is also a problem in the insurance business (how do you get people that will be covered for the consequences of their acts not to behave dangerously?) and it is usually accompanied by a problem called "adverse selection" (if you make loans more expensive to reduce the moral hazard problem, only the people who expect not to repay you will be motivated enough to borrow...).
It's not easy to find the right balance between making banks paying for their sins and protecting the real economy from what happens if a bank folds; banking regulators have therefore tried to impose on banks more and more detailed rules on what they can do and cannot do, and, more to the point, how much capital they should hold to do every kind of operation. This led in 1988 to the creation of the Cooke ratio, which basically says that you must put aside as capital 8% of the money you lend (you can lend any "real" money you own no more than 12.5 times). A few exceptions were created for some safer categories of loans, but the system was quite primitive. There is a massive exercise currently under way to improve these "capital adequacy requirements", usually known under the name "Basle II" (because both the 1988 and the current exercise were run under the aegis of the Bank for International Settlements, the Central Banks' Central Bank, which is based in Basle, Switzerland.) It is horribly complicated, so I won't go into more detail here; but the ironic thing is that the most sophisticated banks will be allowed to decide on their own how much capital they will put aside for each operation they do (of course, they will have to follow consistent rules that are supposed to be monitored by the regulators, but essentially, they will make their own allocations). Back to square one and the Venise merchant bankers trading on their good name alone (with a big bureaucracy attached to them)...
3. Anyway, that's the theoretical (and macro) part. let's now go into the practical (and individual) ways in which bankers can be dangerous...
- commercial bankers see their remuneration based more and more on "upfront fees", i.e. amounts paid by the client on the date of signing of the loan (or the date of disbursement) in addition to the interest which will be paid throughout the duration of the loan. Future revenues are thus of little interest to the bankers deciding how to price a transactions. Client can be tempted to ask for longer durations, accept a larger upfront fee in retunr for a lower interest rate afterwards. They are happy, the commercial banker is happy, and the bank gets stuck for a long time with a poorly remunerated risk.
- commercial bankers are not really blamed if they are wrong with everybody else. They are victims of "the Asian crisis", the dot-com bubble", "the real estate crash", and, as long as they did not lose more money than their counterparts in other banks, it's not too bad. Of course, they all went there in the first place because it looked attractive and juicy, and after a while, the money lent did not support viable propositions. Bankers work in herds. Once one has shown the way to make money, they all pile in; as there are more of them, they start offering better and better terms to the clients, take more risk for less reward. "It's the market" they say when they are in, and when it crashes, they only did like the rest of the market, so they cannot be blamed... On the other hand, stay out of the market when the going is good and you will be blamed for not bringing in revenues like the others. Very few banks, in my experience, seem to look at the total returns through the good and the bad times. And worst of all, commercial bankers in many financial sectors are recycled every few years, so those that might have the experience of what not to do are gone, leaving only new guys keen to make their mark, make a quick buck and fall in the same pitfalls (or new ones) on their own.
- the safeguards against these behaviors is that commercial bankers do not decide whether to make the loan or not. A "Credit Committee" decides this, and it is usually composed of the top management of the bank, with advice from the "credit department", or "risk department", i.e. the guys playing defense, who are supposed to know what can go wrong in a deal and kill unreasonable propositions. Commercial guys thus spend a lot of time "selling" their deals to the risk guys, finding a compromise on what can be acceptable. When it works well, and the risk people are competent, the results are reasonable for all, but the system can easily go astray.
The top management can care more about generating income, league table status (that's the rankings of banks by the number of deals or the value of deals they have done) or imporving their stock price by getting short term income quick; if they are powerful enough, they can overrule the risk guys in the decision-making process. Depending on how the risk department is structured, it may not know as much on the market as the commercial guys and may not always be in a position to argue against deals with enough information. Risk people do not get bonuses for deals, ans may not have the same motivation to fight a dela than the commericla guys have to conclude it. This is an organisational issue which has no simple answers, and which various banks handle more or less well.
- banks rely a lot on "syndications": the bank that makes a loan shares the risk with other banks. Some may say: "oh, if X has done the deal, it must be good, so let's go in" without really checking what's in it. The commercial people will say "this is a popular deal, we have to get in or we won't be credible in the market afterwards", "it's a chance to get closer to that client", etc... This reinforces herd behavior, fads and self fulfilling prophecies (a deal is popular because it is popular) and can lead to bubbles and reckless lending. Now, everybody is supposed to check everything, but this is not always done, and you'd be surprised at some of the things that go through the scrutiny of many extremely qualified people in many respectyable institutions without being noticed...
- in the bond market, it is even worse, because the banks that structure the deal (define its parameters and negotiate them with the borrower) do not even keep the risk - they sell it all to the bond market. Once it's sold, they don't really care how it fares (especially as the restructuring teams are usually completely different form the initial structuring teams if there are any problems). and how do they sell their bonds? By having massive sales desks who work, for a good part, on their reputation ("if it's X selling the bond, it' s surely good paper" thinks many a bond buyer). Now a reputation is something worth defending, but we see that in heady times, things can go astray pretty quickly.
Luckily, we have not had a banking crisis in recent years, thanks to decent risk management by the banks and a long streak of increasing bubbles which have each saved banks from the bursting of the previous one. The LTCM debacle in 1998 brought us really close to a total worldwide financial meltdown. It's a bit more complicated than that, but they essentially played a supposedly safe game as described at the top of the post, and forgot that they could run out of money (in their case, they made massive bets and forgot that these bets made sense if they had a counterparty - when the markets dried up because of the nervousness from the Asian and Russian crises, they got stuck with huge open positions that they could not cover). Who know what the newt one will be?
Anyway, just because a banking crisis has not happened recently does not mean that it won't happen again. Bankers are really smart at finding new ways to lose money, especially when it's not theirs.
If you are still wondering what the link is between my title and the first part of the text, there isn't, really. Maybe that - there is no free lunch. Risk has a funny way to come and bite you when you don't expect it - and you should never let the people who have/create/play with you money forget that.