The effect of Basel 3 and other bank regs...Winners and Losers

The net effect of all these regs to discourage financial innovation, increase capital and liquidity costs, encourage consolidation based on scale economies, discourage moves to universal banking based decreased scope economies, encourage moves to lower tax regimes (mainly in the east), discourage securitization, discourage otc derivs, encourage exchanges and CCPs, increase risk management especially CCR and MR, encourage collateralization, make trading books smaller, banking books bigger, increase demand for ops efficiency,increase reg arbitrage, increase compliance and reporting costs, discourage alternative assets.

Moral Hazard is a wonderful thing! NOT.

Moral Hazard is a wonderful thing! Part of the fundamental role of regulatiom is to avoid it, and yet in our increasingly short term world, politicians and voters alike are unwilling to take short term pain for a long term benefit. Like letting some of these companies pay the price of poor liquidity risk management. Even liquidity/capital injections just make the problem more trenchant the next time. Sometimes i think that risk management only works against a stable context, some lender or provider of liquidity of last resort, a government, an insurance company, and IMF, and without it the network of relationships crumbles. As institutions get bigger and more global, there is no longer some JPMorgan in the background ready to bail out the system. Perhaps this suggests too big to fail really is just too big...

Proprietary trading at federally insured banks

I have never really bought into the need for proprietary trading at federally insured banks. The ugly undisclosed truth is that most banks have no competitive advantage in much of their trading activities, if these operations were evaluated on a risk adjusted basis, they would be, and have been, value destroying. If truly private institutions want to do this that's between them and their shareholders, but when public guarantees are made, then the government has a responsibility to intervene. Moral Hazard is alive and well. Banks like the rest of us, delude themselves into believing short term gains are indicative of internal competence. In a somewhat efficient market, most of us have beta but alpha is and will always be a rarity.

Technocrats vs Politicians Round 1...

The world of bank regulation is being torn between two very different agendas. First, the technocrats of Basel, as they struggle to put in place amendments to the comprehensive version (June 2006) that has formed the bedrock of much market and credit risk management thinking for the past few years. BIS has introduced a host of new consultative documents, subject to ratification, that promise significant increases in the levels of regulatory capital and liquidity particularly for the trading book. Initiatives like the stress var calculation, the calculation of an incremental risk capital charge for users of internal models for specific risk, counterparty credit risk capital increases, the Liquidity Coverage Ratio, the Net Stable Funding Ratio and the Leverage Ratio, all proffer band aids to weaknesses in the original accord. The danger is that the technocrats responding to real issues in the original accord lose the "high ground" and surrender the two principles that made Basel II such an innovation - capital is alligned with risk, and the better the job we do of measuring risk the lower the capital charge. Unfortunately the "band aid" approach understandable as it is, risks making an already highly complex set of regulations, conflicting, opaque and even more open to the regulatory arbitrage that brought down its predecessor - Basel I.

So not surprising therefore is the new found emphasis on the agenda of the second group of actors in this drama - the politicians. They are reacting to market and economic crisis, and selling their approaches to an angry and impatient tax paying public. Here initiatives are basically four fold. In the US, a raft of initiatives have been proposed by the Obama government, in general limiting the scope and the size of federally insured bank activities. The real challenge of all these initiatives is of course congress, where well funded lobbyists may easily derail the worst vicissitude of the proposals. The UK has taken a different approach, experimenting with macro prudential regulation an structural regulatory change, but they face two different challenges - increasing regulation yet not killing the Golden Goose of the City that sustains a sizable chunk of the UK economy. Added to which is the upcoming election which may endanger any initiative. The EU has been hamstrung by government deficits in certain member countries and has revealed just how hard it is get cross the board agreement in such national sensitive economic issues. Finally enter the G20 and the IMF, frustrated at the lack of regulatory changes a full year after the crisis, has argued for the simplest, but possibly most destructive approach of all, bank taxes based on the size of liabilities.

So whose constraints will determine the future of banking for the next ten years? I vote for the politicians - but it will further push financial innovation to the east and away from established financial centers.

On having multiple discount functions

Another interesting development in the world of Behavioural Finance is the idea that individuals have multiple discount functions. For example, researchers such as Prof David Laibson argue that we have at least 2 discount functions, one for the very short term say minutes, hours, days, and then another for the longer term. The tension the investor faces is the tradeoff between these two. For example, simple exponential discount functions argue that the change in  discount rate is more or less constant over time. Yet consider this thought experiment - imagine you were given two sets of choices 1) a) $100 now or b) $101 in 60 mins from now and 2) a) $100 in 1 week or b) $101 in 1 week and 60 mins. Almost everyone will choose 1a and 2b but this is not consistent with a simple discount function. The attraction of the immediate is overwhelming for most people, percieved as far more beneficial than any delayed extra value. Yet beyond the immediate people switch to a more rational process incorporating the extra delayed benefit. Its as if they have two discount functions, one that kicks in for the near term, emphasizing short term benefits and the second which is used to evaluate longer term costs and benefits. Even more fascinating some researchers are arguing that these discount functions actually correspond to different more or less developed parts of the cerebral cortex, associated with high level reasoning and more emotional animalistic instincts. But these discount functions are not merely quantitatively different, they seem to be qualitatively different too. For example, when you change the reward, say to ice cream or a candy bar, the short term benefits even more massively outweigh longer term benefits.

Of course, marketing people instinctively know these things we suspect, the offer of the free ipod when you sign up for a new credit card, peoples tendancy to spend hours negotiating on physical aspects of a car purchase, and then only a moment to decide whether to get auto financing (which is where most of the profits are made).

So what is the punchline - real physical things obtained immediately are valued much MUCH more than abstract non physical things delayed.

What do probabilities mean to people? The Cost of Certainty

One of the most interesting notions in Prospect Theory, is the idea that individuals really don't understand probabilities. That is they make little distinction between similar probabilities (say the difference between 0.25 and 0.3) and treat them essentially the same. However that is not true at the extremes, we seem to have a cognitive bias in favor of zero and 1 as absolute certainties. Yet ironically, absolute certainties are never the case. Mathematically this is described as a weighting function that converts quantitative probabilities into qualitative weights - using a function like the following:

So basically we underestimate probabilities when they are slightly less than 1 and overestimate them if they are slightly more than zero. What does this mean for finance? Well, people will pay more to receive contingent cashflows which are very unlikely because they will overstate the actual probability. This is part of the attraction of lotteries - although the probability is small, there is a small probability of success and that makes all the difference. Similarly investors will heavily discount cashflows that are not absolutely certain.

Prospect Theory and what it means for economics

Traditional Economics is based on the notion of a utility function which increases as a concave function) with increasing levels of wealth. Our decisions as economic actors therefore boil down to selecting the option that generates the highest expected utility. But ask yourself this, faced with a fair one off gamble in which you might make 2000 dollars or you might lose 1000 dollars, what would you do? Most people would choose not to gamble even though the expected utility will be positive. Why? Note that this is not true if we were able to play the game say a thousand times - in which case almost everyone will take the bet.
Behavioral theorists suggest this unwillingness to take the single bet is because we care about losses more than gains, counter to the principles of traditional utility theory, and so have posited a prospect theory that charges a greater value loss for losses in current wealth that for gains. If you like, the utility function is "kinked" around the current level of wealth.

Most of financial markets are based on a zero sum game. Consider a structured product. If I win, you must lose and vice versa. Financial engineering can only add value to such transactions by exploiting the differences in the preferences of the transactors. For example by emphasing the elements of the transaction that most appeal to the other side, say high returns, or delayed payment, or low risk. Being a zero sum game, these elements are paid for by subtracting from aspects of the transaction that the counterparty does not care so much about, like operational complexity, or like risk.  Value is created from the transaction (but not utility however) if the structure reflects the differences in the counterparties wants and needs. Prospect theory  acknowledges that value is inherently client specific (it depends on their reference point for example), and that financial product design is really about understanding behavioral biases in order to better align. So is Prospect theory purely descriptive while Utility theory is normative? Should we give up structured products because they pander to our cognitive weaknesses? Although Prospect theory is partly descriptive, it is also normative. Other counterparties have cognitive biases too, and we must interact with them, so it behooves us to understand and acknowledge these biases/limitations/constraints in our transactions with them.