My Unspeakable Observations On Finance
January 7, 2011 8 Comments
Lots of people have a lot of ideas regarding the financial system, what went wrong, and how it should be fixed. For example, Dodd-Frank, or ‘let’s cap leverage’, or my personal favorite, ‘what if we just banned all derivatives?’
In case it might help advance the discussion, let me lay out some principles of finance/banking that I have observed and amassed:
- The primary endeavor of finance is to try to extract maximum value from capital within the letter of the law. One corollary is that ‘let’s make better regulations’ is never the solution to any problem. Okay, make your regulations, revise them, etc. Big deal. The finance world will read them, analyze them, and then figure out how best to push, stretch, and route around them. That is precisely what they do. I honestly don’t know what the ‘better/more regulation’ dreamers and fetishists think they’re accomplishing. Ever. In fact, more regulations just means that the task of ‘figuring out how to extract value within the regulations’ gets that much more technical and difficult, requiring that much more skill and added-value – in other words, it practically guarantees ever-increasingly obscene compensation in finance. If this is the outcome you seek, lefty, then fine, but don’t kid yourself about what you’re doing.
- (Relatedly) A large and largely-unnoticed effect of ‘regulations’ is to create pointless, zero-productivity jobs for overeducated people who want and demand six-figure incomes doing something, and to guarantee similar lifestyles for their kids. The more regulations you write, the more people will get jobs in ‘compliance’. Or as the ‘chief operating officer’ (which, as far as I can tell, is something like a ‘corporate mommy’). Or ‘reporting’. Or ‘consulting’. This means plenty of jobs for people in cufflinks or pants-suits, having meetings with each other and making Important reports about how the banks are compliant with this and that, or collating numbers together to prove that banks are measuring their risk appropriately, and then submitting their Important findings to the government (to be filed away in Important file cabinets), and then doing it again next quarter, and so on, for a lifetime of reasonably well-paid (or in some cases perhaps even ridiculously-well-paid) pathetic make-work. They will live middle- to upper-middle-class lives, many will send their kids to private schools, and those kids will grow up to be bond traders. If this is the outcome you seek, oh so egalitarian lefties – creating, maintaining, and paying a permanent technocratic overclass that produces no product of any value (not even just pure PnL!) – then fine. But don’t kid yourself that you’re doing it to help out Iowa truck drivers or Louisiana fisherman. You’re effectively doing it to help out your classmates from Brown.
- The supposed bright line between ‘market making’ and ‘prop trading’ is a complete fiction. Often people toss out these ideas to ‘ban prop trading’, or to better wall off prop trading, or ban certain sorts of trades ‘unless they are purely hedges’. This is all nonsense, much of it incomprehensible, as if made up by someone with no knowledge of the subject. To be sure there is this nice polite fiction, or fantasy, among some – and nurtured by the banks themselves obviously – that there is such a thing as a pure market-maker role that takes no risk, and so if we could somehow confine banks to doing only that, all will be well. But market-making without ever taking a position is unrealistic, and there is no logical way I have ever heard anyone articulate of circumscribing what constitutes ‘prop trading’ and what doesn’t (nor can you, in finite time, ever hope to work out which positions constitute ‘hedges’ vs naked positioning – for starters, at which level of aggregation?). If a market maker, seeing customer flows and anticipating the direction of the market, positions himself ahead of the flow, makes PnL from smart positioning, then exits the position a bit later at a profit, was that a ‘prop trade’? Why or why not? Show your work. I honestly don’t know, because it’s a phony distinction. Some (market-maker) traders will tell you that they make most of their PnL from doing what is essentially prop trading (and, which is typically trading the risk/control groups try to prevent them from doing), and the market-making side is small potatoes, if not an outright burden to have to provide liquidity for clients. Taken seriously, this means that what banks get most profit out of doing, they have to pretend (even to the point of hiring teams of people to try to stop it) not to be doing it. This is a schizophrenic situation and it’s not clear why the whole kabuki dance is even necessary.
- The lack of knowledge of would-be regulators/controllers is inevitable and insurmountable. A lot of regulation and thinking about control/management is tacitly premised on a superhuman amount of ground-level knowledge about the detailed trades and positioning at the desk level, knowledge that simply can and will never obtain in the real world. For much the same reason that Communist central-planning fails, the technocratic/regulatory mindset fails: the technocrats don’t and can’t know what they’re doing or talking about. Often/usually not even the head of the actual desk inside the bank itself will have much trade-level detail about what’s going on in one of his trader’s books; the head of a division will only have a vague idea; the head of the investment bank will only know some very gross and aggregate (in some cases, e.g. VaR, mostly meaningless) numbers prepared lovingly for him by others. A recent seemingly-desperate, grasping response to this problem is to just demand that banks throw an ever-increasing amount of data (risk metrics and such) at regulators, in what is either a shotgun approach (‘let’s just ask for everything and maybe that will help’) or an attempt at CYA (‘we couldn’t have possibly asked for more’). What the regulators, or the Fed, or whoever, think they are going to do with all these terabytes of data from all the banks is beyond me. Besides employ a bunch of IT and techie people to try read/interpret them. (See above re: make-work jobs.) Even if they were capable of dealing with it and analyzing it (which they’re not) what would they actually do. The point is that as the financial system gets geometrically complex and globalized,
central plannersregulators naturally have less % of the information they’d need to achieve their aspirations, not more, and this only gets worse over time, not better, and asking for a geometrically-increasing amount of data you can’t possibly do anything with is no solution. Regulatory approaches that don’t take this issue into account are retarded. - The concept/crime of ‘insider trading’ is incoherent and it should just be legalized. A lot of stuff takes place legally every day that to me is logically indistinguishable from ‘insider trading’. This rule and rules like it seem more image than substance; the aim is plainly to instill ‘confidence’ in the market so that people will play in it. So if that’s why it’s needed, then fine, but let’s not kid ourselves that such measures (and other, more Puritanical rules about markets and what their participants can/can’t do) have anything to do with protecting the common person or are in any way ‘progressive’. ‘Confidence’ is of course the etymological root of the word ‘con’. Keeping up appearances so that people stay fooled and the con game can go on is not an egalitarian, populist aim.
I could go on but my opinions/assertions would just get more heterodox and unmentionable from here. The point is, these views of mine, formed by observation, are at such variance with conventional wisdom, such a 180-degrees rotation from what most people think, that for me to follow news, commentary and debate on finance, regulation, and related issues practically causes me physical pain.
The saving grace is that maybe I am wrong about some/all of these things so if you have reason to believe that is the case then feel free to speak up (you’d be doing me a favor), but as of now I’ve seen no reason to think that I am wrong or that the chattering classes know what the hell they’re talking about regarding any of this – unless of course they continue to pursue this retardedly technocratic approach to regulation in full knowledge that they are self-servingly supporting the upper class, their own salaries, and their gifted childrens’ futures as bond traders in doing so.
Agree completely that the conventional wisdom is close to 100% wrong on each of these item.
Disagree on this point:
“One corollary is that ‘let’s make better regulations’ is never the solution to any problem. Okay, make your regulations, revise them, etc. Big deal. The finance world will read them, analyze them, and then figure out how best to push, stretch, and route around them.”
If you regulate activity that’s the result. If, on the other hand, you regulate structure you might actually solve the issue by aligning incentives properly. My pet regulation theory is that you should require all banks to be structured as partnerships. Much harder to make heads I win, tails the treasury bails me out bets when you have personal liability. Who knows? Maybe all the banks move offshore. On the other hand, Goldman Sachs operated as a partnership for almost until the turn of the 21st century.
The idea isn’t that this time the regulators will know enough and this time they’ll have the authority and initiative to stop things before they go wrong but that the management will actually do so because they’ll be penniless if some trader blows up the bank.
You have a point; for the sort of critique I’m trying to make, it’s useful to distinguish between regulations aimed at activity vs. those aimed at structure, and the latter may be less doomed if done right.
I’m skeptical of the partnership idea but can’t immediately say it’d obviously make things worse. I do still hold that whatever regulations are put in place, their boundaries will be explored and if such and such route-around is found (you mention moving offshore), it will be exploited maximally. This will inevitably lead to currently-unforeseen problems and a clamor for yet more ‘solutions’.
On the particular point of incentives, you’re obviously spot-on the the problem is heads-i-win, tails-taxpayers-lose bets. But if I had my druthers I would address this problem by focusing more on simply eliminating (by fiat or even constitutional amendment if necessary) the ‘taxpayers lose’ option than by trying to align incentives. Think about my point regarding information: in your regulatory vision, you can align the partners’ incentives all you want, but they don’t and can’t ever possibly know all the risk they have taken on. The head of the desk itself doesn’t even know, why would the partners know? So their incentives will all be in the right place but this won’t actually prevent blow-ups of the sort we saw in ’07-’08.
The key in my view is just to not have the government bail anyone out on blow-ups. Ever. No AIG bailout, no Bear bailout, no Citi bailout, (no auto bailout, no….etc)
Obviously, this is one of those rare things that would be both politically untenable with the masses and contrary to the reigning philosophy of the technocratic elite, thus it is absolutely not going to happen. Like I said, my views here are 180 degrees away from conventional wisdom. Oh well….
best
“Obviously, this is one of those rare things that would be both politically untenable with the masses and contrary to the reigning philosophy of the technocratic elite, thus it is absolutely not going to happen”
Maybe not so impossible. The bailouts were wildly unpopular among ordinary Americans, and a candidate who ran on a “No More Bailouts for the Rich” ticket would attract a lot of grass-roots liberal support as well as Tea- Party types. If handled properly, I can see this as a winning ticket for someone in 2012.
Were the bailouts all that unpopular? My sense is that the median American basically buys the basic concepts of Keynesianism (‘stimulus’, etc), always favors big government gestures in the face of supposed ‘crises’, and would punish, in polls and at the ballot box, any government that they saw as ‘doing nothing’ in the face of such a ‘crisis’. And however much people may voice opposition to the general principle of no-bailouts, it always seems easy to argue for this or that exception. ‘But the auto workers!’
When the rubber meets the road, then, I don’t see the public as all that opposed to bailouts in practice, not in any way that matters.
Also, while I understand a common way to read the 2010 midterm elections is probably as a repudiation of bailouts (and the health care legislation), I don’t read them that way and I think conservatives/Republicans who do risk a miscalculation.
I could be wrong though, and I would like to be.
“Think about my point regarding information: in your regulatory vision, you can align the partners’ incentives all you want, but they don’t and can’t ever possibly know all the risk they have taken on. The head of the desk itself doesn’t even know, why would the partners know?”
I might be wrong here but I don’t think that trading has to have unquantifiable risk. Right now no one cares about the real level of risk and as a result there is no serious thought put into how to properly measure and evaluate risk. No one today really tries to measure risk – they get no benefit from it. All the benefits accrue whether you actually evaluate risk or just pretend to do so.
Look up and down at the industry; risk management, which should be the cornerstone of a business basically built on risk arbitrage, is treated as a joke. Ratings agencies exist to stamp certain debt as eligible for holding by entities who are legally required to hold agency approved debt. Risk analysis and reporting is ignored make work. Risk departments are thought of as a necessary evil; have to keep them around for the regulators but traders should just ignore them whenever possible. This is all an outgrowth of the fact that management really has nothing at risk.
Right now, if your firm is better at evaluating risk than the competition you get – what exactly? Really you just get smaller bonuses because you passed on investments that other people would buy. Maybe they blow up and maybe they don’t. If they don’t blow up (you’ll win at Russian roulette 5 times out of 6, after all) you just passed up money – lots and lots and lots of money. In a world where financial executives wind up broke if their firm is bad at evaluating risk (and they happen to roll snake eyes) they might actually try to evaluate risk.
Partnerships with personal liability might actually get us closer to that state. They might even be too risk averse.
I might be wrong here but I don’t think that trading has to have unquantifiable risk.
In principle you may be right but in practice, in any practical organization, whatever risk is being quantified will always be an incomplete and inaccurate view of the full risk. (This is the one nugget of agreement I think I may have with Taleb.) The idea that all risk is ‘quantifiable’ (which may be true theoretically), and therefore it’s possible to attain a state where you have quantified ‘all’ risk, seems misguided to me. It’s not merely that you will miss measuring some risks, but the risks you are measuring will have errors in it – always. Not merely first- but also zero-order errors. Maybe those errors are negligible and maybe not, but any regulatory scheme that relies on (by whatever mechanism) getting to a point of full risk measurement is fighting an uphill battle against increasing complexity and seems doomed to me. I know your point is that the hypothetical Goldman partners will be super motivated to ensure the risk is being fully and accurately measured. Aside from creating more jobs for risk management and IT professionals, I don’t think it will help much.
Also it seems wrong to suggest that ‘no’ serious thought is put into measuring and evaluating risk. There is plenty of work done in the field of risk management and so on. I guess you could say that the work isn’t being paid attention to. I might say in response that I’m not sure how much it should be. Because I’m not sure I even agree that the goal should be able to get banks reduce risk per se in the first place. Clearly we want banks taking on risk. Do we want loans? Every loan is a risk. As you say, one outcome of your idea may be that banks become too risk averse. That would not be merely a case of being ‘overly cautious’, that would actively be bad for our economy. So how do we incentivize banks to take on the perfect ‘correct’ amount of risk. I’m not sure we can because I’m not sure anyone knows or can define what that is. And centralized attempts to do so, to prescribe exactly how much risk banks should hold, based on formulas and assumptions of giant amounts of data, will only take us further down this technocratic-regulatory path we’re on (and, of course, create more technocrat jobs at, effectively, taxpayer expense).
You mention rating agencies and they are a good point of the folly of any top-down, centralized, technocratic approach to risk. Why does anyone care what rating agencies say about anything in the first place? Only because their ratings were given the power of law in their role in capital requirements. (edit: And that certain types of buyers are restricted to buying certain ratings.) See this post of mine. In my perfect world it’s not that rating agencies would ‘measure risk better’, it’s that no one would pay attention to them (unless they happened to come up with good and useful findings – by whatever method). Relying on rating agencies to measure risk, getting mad when they fail, and then diagnosing the problem as that failure, is part and parcel of the narrow mindset I would step out of. The problem is the policy that assigned their statements power, not the content of those statements.
Finally there is the idea that the problem is financial executives having no incentive not to play Russian Roulette. But I don’t think that’s the case at all. If their firm fails due to a trade gone bad, generally they will personally lose a lot, because typically a large portion of all such executives’ pay comes in their firm’s stock. Well, I guess they may not pay so much if their firm gets bailed out, but that’s exactly the outcome would like to disallow – by not bailing anyone out!
Perhaps the one thing that could be done to address this problem (if it is the problem, and I’m not convinced that it is) is to change how PnL is measured. I know some people have this idea (because I hear it from non-finance friends) that the problem is that a trader’s trades could make X million a year for a few years, and the trader will get paid, but then in year 6 when he’s gone they lose Y>>X. My response would be that (in such a scenario) in those years 1-5 those trades probably didn’t really ‘make’ X million at all, and the bank’s books and records should not have pretended that they did. This could come down to marking to model, lack of proper provisioning, and so on; the causes may vary, but one could imagine a different framework, wherein PnL is only released over time in accordance with the ‘blowup risk’ rolling off. (Which requires better risk measurement, obviously.) Of course, banks are required to mark to market and there are ‘generally accepted accounting principles’ regarding what their balance sheet should say…all imposed by super-smart technocrats and regulators of course…and this Russian-roulette problem has come to exist within that framework. Maybe that should tell us something?
best
Pingback: dustbury.com » We’re in the wrong business
Pingback: RWCG: The Crib’s Notes Version « Rhymes With Cars & Girls