Wednesday, June 23, 2010
Let's start with a situation once witnessed while working in the Far East markets.
When a junior trader was asked what he had in EUR/USD he said he was small long. Why? Because the trader next to him was longer and he knew something. When asked what he knew, he replied that his Head of Desk was long even more. So the Head of Desk was approached and interrogated. It turned out that he was long because the Head of the Room was longer still. Not easily put off the trail, he too was approached and asked if he knew anything about EUR/USD? "Not really, but the Treasurer has put a position on so something must be up". On we go. As it turned out the Treasurer was long because "A friend at another bank had told him that China were on the bid". When told what the whole room's exposure was on the back of this chain of rumour he was horrified. And this was only in one bank. If this was being repeated in banks across the region then the positions amassed on this whisper were huge and at some point were going to have to be unwound, which duly happened in a massive dump at about 9.30 London as Asia went home.
So why does this happen? Well behaviourally it's simple really. If you mirror your boss's position in a smaller amount you have 2 possible outcomes:
- You make money.
- You lose money, but not as much as your boss so he can't chastise you.
Since then watching the Asian close has always been an interesting time of the day to gauge the spread between rumour and fact. Which brings us to today. The Chinese-inflicted kicking of the FX market yesterday appears to have elevated their general FX status back to old Voldermort levels. And it appears that the whole world is so terrified that they may be buying EUR/USD around the 1.2240/50 level that is enough for everyone else to be long ahead of that level. But how much?
It could be ugly if the pit-prop isn't as strong as hoped.
But this perceived "Win/Not lose" payout can be seen all over the place. Around dinner tables house prices are too often discussed. Either money has been made or in a falling market they are crowing about how they haven't lost as much as others. Benchmarks are effectively chosen to suit the message that has to be sold. And benchmarks are mostly bollox. They are a crutch with which to abdicate blame. Equity managers down 20% crow about 50 bp out-performance to benchmark, effectively pinning the blame back on the poor mug who ever decided to allocate money to them in the first place. And considering that a market is the sum of all players and the players are the fund managers then net performance can only be flat across the sector. So why waste massive fees and not just go and buy the index?
The application of benchmarks is too easily assumed to be a way of strapping down risk and measuring performance. But strict rules of process end up applying step functions to a world which is actually governed by curves of risk. For example, restrictions placed by trustees on investment institutions to buy "only investment grade" means that you can end up with two bits of paper which could have a micron of separation between their real credit worthiness but be just the other side of the ratings agencies "cross-over" and, as during the credit blow out, be priced 1000 bp differently. And a real world example: why is it that just within a 30mph zone you are a mass murdering lunatic to even think of doing 31mph but 1 inch beyond the restriction 60mph is just fine? And woe betide you if you are ever caught by the police doing 1 mph over the speed limit and, after being lectured on the dangers of speeding, you try and argue the REAL risk differential between 60mph and 61mph....
The problems really kick off when the assessment of reality of risk is divided up between different departments or units. Imagine if, using the driving analogy, someone in the foot well next to the accelerator was responsible for speed risk, someone else tucked under a wheel arch was responsible for directional risk, and someone strapped to the front responsible for impact risk all reporting back to a blind bloke in the driving seat. Going to be a crash isn't there? But this has been known to happen in the real world too.
Another story starts with a chap from a bank's product control department visiting a trading desk and asking them why they are losing money on client trades. The trader replies that the head of sales had told him that the desk needs to be aggressive in its pricing so that business can be won for the bank in "more-profitable" areas, such as exotics. At this point, the risk manager remembers that one of the exotics desks he is responsible for has just started posting some very large losses which he was going to investigate that very afternoon. So he goes to the exotics desk and asks the trader why they are losing all this money, who replies because of xyz there was a big correlation breakdown, the markets jumped and there was no way to hedge the risk. The risk guy replies "But I've been sending you these reports each day with all your correlation risk etc. Why didn't you take this into account?". The exotics guy replies "well, the head of exotics sales told us we had to be aggressive in our pricing to win business and warehouse the risk, the head of trading was fine with it". So the risk guy goes to find the head of trading and asks him why he was fine with them having all this risk? The head of trading replies "huh?". After a brief conversation it turns out that the head of trading doesn't understand the risk the exotics desk were running and thought everything was OK. Net result: flow desk loses money, exotics desk loses money, both flow sales & exotic sales get a big bonus.
Now where is this all leading us? Don't know really, but we do know that there are stupid things happening all over the place because of righteous laws applied in an unrighteous world