Friday, January 22, 2010

What is proprietary trading?

Somewhere in the debate about prohibiting proprietary trading in certain banks we will need a decent understanding of what proprietary trading is.  So I thought I would illustrate the difficulties.

Imagine a suburban bank which takes deposits and makes mortgages. 

The deposits are primarily at-call and pay a floating interest rate.  Legally they bank has overnight money – and if interest rates rose then the next day the customers could (in theory) all withdraw their money and/or ask for a higher interest rate.  The bank does not really know what interest rates it will be paying next week let alone in three years. 

In reality the customers of the bank are sticky.  There is no way that everyone will pull their money in response to a short term rate rise.  The funding of the bank is of uncertain duration

On the asset side the bank lends on fixed rate but refinanceable mortgages.  The bank really has no idea how long the mortgages will last.  If rates go down they might all be refinanced tomorrow.  People might just sell all their houses and repay their mortgages.  In reality however the customer are likely to be somewhat sticky.  On the asset side the bank has uncertain duration.

This plain vanilla bank has interest rate risk.  If rates rise their funding costs will rise relative to their asset yield.  If rates fall their assets will refinance.  Their funding cost might also fall – but at the moment the funding cost seems pinned by the zero-bound. 

Some hedging of interest rate risk here seems entirely sensible.  Banks (and more often S&Ls) have failed in the past because they failed to hedge this sort of interest rate risk.  However as both the assets and liabilities are of uncertain duration there is no way of knowing just how much hedging is required.  There is a choice here – it is a proprietary choice (in that the bank will trade off hedging costs against profits).  And there is no easy way to legislate that choice away.

I have even seen a bank swap its fixed rate subordinate funding (fixed rate preferred shares) into floating rate and call it a hedge.  That looked like proprietary trading to me as it earned a profit (the yield curve was steep) and the bank was playing the yield curve very heavily in advance of that hedge.  The so called hedge increased the risks the banks faced if short rates rose.  The auditor signed it off as a hedge.

If the auditor, the bank and I disagree as to what is a hedge in the simplest of examples then I have no idea how we are going to find a legislative solution in complex examples.

Real prop trading is like pornography.  I know it when I see it.

 

 

 

John

22 comments:

  1. Though this analysis is true in theory, I would think that in practice the concern is not the banks who "play around" with what is or isn't a "hedge" or uses "trading" to reduce their costs of funds, but rather those who employ people for the explicit purpose of making frequent, highly leveraged bets on small market fluctuation.

    Given that the later activity is (relatively) easy to distinguish, it would also be (relatively) easy to reduce.

    Regulation would not eliminate the risk ... but it would reduce it.

    SamB

    ReplyDelete
  2. I can't tell frequent transactions for pennies from market making. Market making is clearly customer driven.

    More to the point - an exposure like Jerome Kerviel (the SocGen trader with 35 billion in equity exposure) is the sort of thing that can take down a bank.

    Someone with low limits and very frequent trades - often done by a computer - looks like it is not going to take down anyone and is very difficult to distinguish from market making.

    John

    ReplyDelete
  3. The government is going to have a lot of difficulties defining the concept of prop trading:
    - A car dealer doesn't sell a car to customer A because he thinks another customer B will buy it for $ 100 more. Is that a bet? Is that a prop trade?
    - A market maker buys a portfolio of bonds to sell them later. He doesn't manage to sell them at the price he likes, and sells them after three months, at a higher price. Is that a bet? Is that a prop trade?
    - You bought CDS on France back in 2006 because it was cheap at 7 bps, and their bond spread over Germany was 12bps. You didn't know what to do with it otherwise, it just looked cheap. And you did a lot of business with French companies, and you relied on the continuing business with French companies. That business has since stopped, but you kept the CDS. Is that a bet? Is that a prop trade?
    - You don't make markets in car manufacturers stock, but you do in car insurance companies stock. Is that a bet? Is that a prop trade?

    The general public has a complete misunderstanding of the market making businesses in US banks. The way trading businesses make money in US banks is no different than the way a car dealer, a carpet seller or a jeweler generate profits. It's just a different asset. Try walking back in to sell the Rolex you just bought. Or the Persian rug. Or the Toyota Prius. Buying and selling. The dealer always has the edge. You don't like it? Get a bike instead of a car.

    Do you want the car dealer to not buy back your old car at all? Oh yes, and you're not allowed to sell it on Ebay either.

    Try cashing out your 401k without a bank's bid on your mutual funds (or its underlying assets).

    I do understand the government's effort to curtail bank's risk taking. But half-hearted and ill-defined efforts won't do. The impact of such actions will be non-trivial, with a risk of unintended consequences. There's really no need to pamper the banks; but simply stopping their ability to take and warehouse risk is debilitating to our money-based economy.

    ReplyDelete
  4. Any any one of two hundred other transactions...

    You do a big equity underwriting. To hedge your risk you go short the market because you can't go short the specific.

    How about guaranteeing a customer VWAP on a trade. Is that prop risk?

    J

    ReplyDelete
  5. undertaking a hedge to minimise risk/ exposure due to agency business is completely different to taking/opening positions with intent to proift

    ReplyDelete
  6. Funding mortgages with overnight deposits is the original sin here. If one wants to do mortgages, one should have to get matched long term funding. If the customer wants to payback his mortgage, he must do it at the market price of the refinancing instrument. That is what happens in Germany. Enough of free options mispriced by financial institutions / Insurances companies and backstopped by the sovereign ! Vanilla is the future.

    As to know what proprietary trading is, it is a simple thought experiment : instead of being on the balance sheet of your employer, imagine that the operation is on your own account, backed by your house and your retirement savings. If you sweat at this prospect, it is proprietary trading !

    ReplyDelete
  7. Every example you've cited is standard asset-liability management in commercial banks. They delta hedge mortgage prepayment risk, they assign core deposits to maturity buckets, and they routinely swap fixed debt into floating in order to match with floating rate assets. None of it has anything to do with prop trading.

    ReplyDelete
  8. Obama's plans on prop trading are pretty comprehensive, and include yet more reform to derivatives. Although reforms to risk management are needed, I worry about the knock on impacts that this will have to world markets, risk management via derivatives, and financial stability. http://bit.ly/8bt4Wx

    ReplyDelete
  9. The best regulation is nebulously-written but has the support of the parties being regulated--you really need the banks to regulate themselves, but with a set of guidelines that allows them the cover to not take on (ostensibly) advantageous trades. Stepping outside of the bounds must be punished ruthlessly--almost without regard for the particulars of the case.

    It probably isn't fair, but it will keep order.

    And you need a chief regulator who, as you put it, knows prop trading when he/she sees it.

    ReplyDelete
  10. Hi John:

    At the end of the day the big banks will get around the restrictions and call it a day.

    In fact from what I've read very few operations at the big banks would fall under the axe, so life goes on.

    ReplyDelete
  11. these are all great questions, John. I think the Administration was going after the concept of the segregated prop desks - most banks have them, and most banks also have what could be called internal hedge funds. however, as you've illustrated, the real question is what will happen to the flow-related desks who take large principal positions.

    Goldman, for example, makes most of it's money "Trading" yet claims a small percentage is "proprietary trading."

    will equity traders be able to trade index reweights? how about merger arb? who knows...

    that's why populist policy reactions usually don't work as intended

    ReplyDelete
  12. It sounds like going back to something like Glass-Steagall would cut out a lot of the grey areas, because the things that look closest to prop trading tend to be related to market-making, which I believe was considered an investment banking activity which had to remain separate from commercial banking under Glass-Steagall. Is this proposed measure trying to do kind of a halfway Glass-Steagall, and that's where the difficulties arise?

    ReplyDelete
  13. Serious treatment of the question, with a memorable an very true conclusion.

    It seems to me the reform proposals were mostly political theater intended to draw attention away from the battering the attention is taking over the Mass special election, the healthcare debate, national security, and so on. If that's so, then the administration may or may not circle back to consider those substantial issues some day.

    ReplyDelete
  14. John,

    I'll admit that I'm not as much of an expert on how the major banks operate as you would be, but I'm pretty sure that already distinguish between market making or hedging activities, and speculative trading on their own account ... if for no other reason than for control and compensation purposes. The fact that from the outside it is not readily obvious is not a bar to regulation. (See also feedproxy.google.com/~r/economicsofcontempt/~3/mwILpjsMSQ0/few-more-assorted-thoughts-on-financial.html) .

    Will some organizations establish complex structures to get around the rules ... sure. But the aim is not to eliminate the activity, reducing is sufficient.

    The rogue trader argument is not on point here ... sure a rogue (or malicious) agent can bring down A bank ... the concern here though is not with A bank but rather with it the banking system as a whole, and it would be pretty hard for one rogue to get replicated.

    SamB

    SamB

    ReplyDelete
  15. Marketing making is proprietary trading, John, full stop. You have to assume principal risk when facilitating customer trades, except in those relatively rare and lucky instances when you can just cross a block of something from one client to another.

    More to the point, even well controlled market making (low book position limits, daily rollovers, etc.) can cause severe losses if market liquidity freezes up. From one point of view, market makers are systemically short liquidity, and when liquidity becomes dear--like in a market disruption or crisis--they can be caught very short and very wrong. I think the only way to limit the damage is to set overall position limits and minimum equity levels for market makers, too.

    Finally, even if you could "immunize" banks and investment banks from all major proprietary trading risk, what about prime brokerage? If the markets go to hell in a handbasket, banks will get hit to the extent they have lent money to hedge funds to trade (e.g., LTCM). This can be a big deal, too.

    I can see no obvious way to limit total proprietary trading risk in the global financial system. If we cannot, then we must accept that banks will always be at risk, even if they are not doing the trading themselves.

    ReplyDelete
  16. The Australian accounting standards, and the international accounting standards on which they are based, now treat all trades as mark to value and straight to profit and loss - unless tough requirements for deferred hedge treatment are met.
    Those requirements? You have to have identified the risk you are hedging; necessarily, your hedge is opposite to the risk and no larger than the risk; and your hedge correlates within specified degrees of freedom to the risk (so that it returns between 80 and 120 where the risk is 100).
    An auditor who signed up to the described transaction as a hedge might do so in the sense that it is called a hedge but marked to value and straight to profit and loss.
    But applying the accounting standards would mean that deferral could not apply - unless it's USA rather than international/Australian standards that apply.
    I don't think the described transaction passes USA deferral standards either, but don't work with them much and would defer to a more expert view.
    Some recent Australian failures depended on more equivocal former accounting standards, deferring and so not disclosing where current standards would prevent this.
    This meant that failed listed companies were not hedging contrary to another business risk (they were multiplying the risk, effectively leveraging their bet), were not limiting the hedge to the amount of the hedged risk or less (the 'hedge' was itself a much larger exposure), using transactions that correlated very badly with the risk supposedly hedged, or otherwise not hedging in the sense of limiting or countering a risk.
    Where banks are concerned, though, it's hard to get the standards applied in a disinterested way and so it's hard to have them really applied.

    ReplyDelete
  17. John, I think that the bright lines here are the liquidity and duration of exposures: if you "own" a market (Goldman SSG in some restructurings) that's obviously prop. All private equity is prop. Having a large exposure for an extended period of time in a single name equity or credit equal to many days' volume is prop.

    I think its very important to recognize that while there will always be some risk taking shenanigans at banks much as there are at big energy companies and power companies, the point here is to make sure the sorts of businesses that add a lot of risk in terms of directional exposure and asset/liability mismatches are kept to a minimum.

    ReplyDelete
  18. Has someone been reading this blog? From Volcker's Senate testimony today:

    Asked by Shelby how excessive bank activity would be recognized by regulators, Volcker quips, "It’s like pornography: you know it when you see it."

    ReplyDelete
  19. The answer to the question as to whether I have Washington readers -

    YES.

    J

    ReplyDelete
  20. Off Topic, but I remember reading an article you did on Spanish Bank BBVA. Did you ever short it?

    http://ftalphaville.ft.com/blog/2010/02/03/140216/bbva-an-exercise-in-spanish-banking-losses/

    ReplyDelete
  21. Indeed, the problem stems from greed, but two-fold or on both sides: ever since courts upheld the (erroneous) opinion that an "overnight" bank deposit was a loan like an interest-bearing savings account was, and thus could be lent out, ever since then bank runs made those who came "too late" discover that coffers were empty, and even IF the bank had not overdrawn, so to speak, it could not get all outstanding loans back to their subsidiaries' cash counters in time due to maturity mismatching. So why would it happen despite the risk? Because otherwise people would have to PAY (more) for their checking accounts. And they are ready to dish out lots of money tom their verterinariansfor their favourite pet, but wouldn't give money to those who guarded their money. Thus lending out a substantial portion (not all, by the way) of deposits subsidises thesechecking accounts = greed on the side of the customer. Then comes deposit "insurance" and more moral hazard, and the fuse is lit after all. And this unsound financing in the long run then drags us all down with it. Were it a free market with prudent investors there would arguably be TWO kinds of overnight accounts: those where the creditor paid a fee for safe-keeping and those where the less risk-avert would instead let the bank make loans with their money.

    ReplyDelete
  22. I guess there is a material difference between the prop trading done by a bank, and the prop trading undertaken by a standalone speculative proprietary trading firm

    ReplyDelete