IT Could Have Caught $2 Billion Rogue Trader

superapecommando writes "With the benefit of hindsight, IT experts are claiming that technical countermeasures at Swiss bank UBS could have stopped rogue trader Kweku Adoboli running up a $2 billion loss." If American Express and Visa can mine transaction data and put a stop order on credit cards when you unexpectedly buy gas out of state, it seems like there could be patterns to watch for when the amounts are in the billions, too.

Comments

perhaps, perhaps not

[Trepidity]

A problem is that it's difficult to design a system to automatically determine "risky" or "rogue" or "non-normal" behavior in investment banking, because taking crazy bets is what they do, and interpreting their official policies loosely is a big part of that. Sometimes it turns out massively well, in which case bonuses all around; other times very badly, in which case start looking for ways to label the guy "rogue" and fire him. But it's de facto, if not officially, part of "normal" operation of an investment bank; it just sometimes turns out badly.

Re:perhaps, perhaps not

[Pinky's Brain]

We shouldn't shut it down, but we should disallow them to use government guaranteed deposits for it. Deposit banks should not be allowed to use their depositor money for leveraged investments or derivatives.

Re:perhaps, perhaps not

[khallow]

It's all relative. That depends on how much they're trading and the tolerance for loss of the people who have the money. If I have a trillion dollars and a high tolerance for loss (and trading with that in mind), then two billion dollar in losses would just be a slightly below average day.

Re:perhaps, perhaps not

[jrminter]

Wish I had mod points... You got that exactly right. Here in the US, the politicians could not resist the clamor to repeal the Glass–Steagall Act of 1933 that made that distinction between commercial and investment banks. Set us on the road to the mess we have now...

Re:perhaps, perhaps not

[kurt555gs]

It is time to bring back both the Glass–Steagall Act of 1933 and the Smoot - Hawley Trade and Tariff act. Unless you really want Des Moines Iowa to look just like the slums of Brazil.

Re:perhaps, perhaps not

[Trepidity]

I don't think Wall Street really believes in efficient markets. A large part of what investment banks do (on the "prop trading" side) is attempting to exploit market inefficiencies via various kinds of arbitrage plays.

Re:perhaps, perhaps not

[TarPitt]

And the culture in these organizations celebrate risk-taking, and handsomely reward those who take huge risks, and will do anything to resist restricting those "heroes".

No organization like this is going to restrict their "heroes" and money-makers with automated software that tries to second guess trading patterns.

I'm saying this as someone who was asked to leave the HR VP's office of one of these organizations for suggesting that compliance with security policies be part of annual performance review. Quoting from memory, "There is no way we are going to penalize our star salespeople for not following a password policy"

BTW, our security assessment found that bad passwords were a huge problem, to the point where a few minutes of guessing gave us some very powerful access.

Re:perhaps, perhaps not

[CharlyFoxtrot]

I think people do relish the opportunity to endulge in schadenfreude where banks are concerned though. People latch on to a phrase like "casino banking" and gleefully apply to everyone in the industry with the intent that we should shut it all down. By the same logic:

A police officer took a bribe, we should shut down the police.
Well, there was teacher who was a paedophile once. We should shut down schools too.
Look at that corrupt politcian, shut down the government while you're at it.
The guy at Blockbuster charged me for bringing a DVD back 5 minutes late. Shut down them too!

The difference with the banks is that in the banks it is a structural problem with the behavior being so deeply ingrained in the culture it's difficult to see how it could be changed. Especially since they have the power to avoid outside pressure to reform.

You do have cases where the police have a structural corruption problem too, or where the government is corrupt to the point where it no longer functions. Just look at many third world nations. And yes, when that happens people need to intervene.

Re:perhaps, perhaps not

[CharlyFoxtrot]

That's the theory and yet all these major banks keep coming out with these rogue traders and that's only the losses big enough that the general public hears about them. Let's face out out on the terrain no-one is holding these guys accountable. IT may set up the system, Risk Management may generate the reports and they'll be either modified to say what management wants to say or just plain ignored because like all gamblers these guys think they have a system which lets them keep on winning even as they are betting their house (or in this case our houses.)

The computer is always guilty

[PolygamousRanchKid]

When a financial boo boo occurs, and IT is involved, it's IT's fault.

When a financial boo boo occurs, and IT is not involved, it's IT's fault.

Computers have a tough time defending themselves, so it is easy to pin the blame on them.

Maybe Watson, IBM's Jeopardy champ, could handle this:

"What . . . is a 'Scapegoat'?"

lack of liability

[azalin]

In my humble opinion the attitude in the finance sector has to change away from gambling and back to investing.
Btw. How comes those people only get bonuses and not fines to?

Re:lack of liability

[prefec2]

Because that would be fair. The whole thing is to privatize profits and let the public pay the bills. And by the way there is too much money on the market to really invest all that money and get something in return. But it is not only the financial market. You can found companies and relay responsibility for bad business decision to the banks, stakeholders etc. To make things right, the size of organizations in the financial and business area have to be limited.

You think?

[igreaterthanu]

Any entrance level programmer could make a system that alerts a superior of the performance of one of these people.

Say an alert when the losses are over $100,000 then over $1mil, $10mil, $100mil, etc. or maybe after 50 losses in a row.

It says a lot about this bank that this wasn't caught earlier. An alert should have been triggered far before the losses got to the 2bn mark.

Re:You think?

[tqft]

The problem isn't detecting the loss after the event the problem is predicting 23 sigma moves.

http://www.zerohedge.com/contributed/ubs-and-big-trade
The trade was profitable until the SNB capped the CHF , capping the trade and trader in the process. At least that is a reasonable working hypothesis. Everything was fine until a week ago. What happened then? The Swiss National Bank announced a cap of their currency vs the Euro.

Why did he have such a large position ? That Risk & Management MUST have known about. And were OK with while it was a winner. When it became a loss, the trader was shafted.

Re:You think?

[julesh]

The problem isn't detecting the loss after the event the problem is predicting 23 sigma moves. [...] Why did he have such a large position ? That Risk & Management MUST have known about.

A few things occur: if it really was a 23 sigma move (I haven't been watching the markets lately, so didn't know what was going on), then the fault really is neither the trader's nor particularly the management of the bank's, but rather the entire industry's. We're talking about an international system of de-facto standard risk management that assumes the worst net losses these traders will sustain is around 5 * sigma. If he'd lost what the models predict he was risking, which is to say about $400 million, I doubt anyone would have made this much fuss. He may have been fired (or perhaps he just wouldn't be getting his bonus this year), but we wouldn't have heard about it. They simply need to start thinking more about the non-trivial risk of unexpected large scale movements (which are not readily accounted for by the standard statistical approaches that are commonly used) and the effects they can have.

Re:You think?

[Anonymous Coward]

if it really was a 23 sigma move

It can only have been a "23 sigma" event if the model by which such statistics are established was utterly irrelevant.

This is the systemic problem with banking... the "risk control" is quite sophisticated - but it takes a narrow view of what constitutes risk... and, hence, no-one should be surprised when real-world risks are spectacularly under estimated. A cynic might argue that the whole subject of risk management is an elaborate illusion to give the impression that risks are being taken seriously. The reality is that where risks can't be argued to be marginalised by elaborate argument - they're outright ignored... and the complexity of the systems is relied upon to prevent scrutiny.

Writing software gives something of a unique perspective on risks... software engineers aim to "prove" the software they write - and bugs are declared whenever any sequence of events is established that would, theoretically, lead to an unacceptable outcome. It's considered "too cowboy" to argue that a bug doesn't exist because the client hasn't get generated that sequence of events in the live system. Conversely, this is exactly how risk management works for financial risks - and it is this deceit that has provided the illusion of large profits and the bonuses that accompanied them.

Re:You think?

[NoOneInParticular]

You do realize that a 23 sigma event implies the belief in a risk model where an event like this will happen once every 10^112 years?!! You also do realize that the entire financial sector is using these risk models, and are therefore still assuming that 23 sigma events will only happen once every googol lifetimes of the universe? A hedge like this, that will only fail once every googol universes as predicted by economics (and is thus a safe bet on the surface), tends to fail every few years.

Sigma measurements of risk is intellectual fraud, on a scale that is costing us billions.

Re:You think?

[NoOneInParticular]

I'm arguing against people equating risk with sigma, and in particular those who say that there's a problem predicting 23 sigma moves, as if that's saying something about the likelihood of the event occuring. I'm also arguing against the financial sector who is still using sigma (and beta, and the entire apparatus of mutlivariate statistics) as something that can be used in practice. You are arguing that the financial sector has moved on, but have they? Are they truly beyond the Gaussian assumption, have they really moved away from Value at Risk models, is Black-Scholes truly abandoned, are standard risk-of-ruin calculations abandoned, or are they still trying to fix Gaussian models with martingale and jump dynamics as they were doing last time I looked into it. What has replaced the entire apparatus that wreaked havoc 5 years ago? I still see bailouts on the horizon, banks very highly leveraged, and a sector that has reasoned away risk using the same flawed arguments that have been en vogue for the past 40-50 years.

You are seriously arguing that this has changed? I've worked at hedge funds, I am working in risk management, the assumption of normality is the only game in town. And yes, I am arguing that all those extremely smart mathematicians and physicists are missing the obvious, simply because they are doing what they were hired for: to create models. Stating that we don't have the mathematics to do the job will not land you a position as a quant. They'll hire someone that will try the best they can: steer a car by rearview mirror, driving in the ravine at the next haircut. All we do is drive a little less fast.

I guess that on the trade floor there might be a few that get it and are using algorithms that do assume that they can lose everything at any time and for instance only trade in options, but at the macro level, capital requirements are still stated in risk weighted assets. Options are still being sold. Individuals and banks are still going naked (leveraged) long. Risk is still assumed to be log-normal. Textbooks are still printed stating this as a fact (not as an assumption). Have you seen a shift in teaching economics, away from multivariate stats? No, as nothing has changed.

Re:You think?

[teknomad]

A 23-sigma event can become a 1-sigma event simply by changing the probability distribution your models use. The problem isn't really the math...sigma measurements of risk *DO* work, it isn't intellectual fraud. The problem is that the banking industry's decision-makers are willfully misusing the math because they do not want to accept what the proper models tell them. It is extremely, extremely, enticing to keep using a Gaussian distribution in your big snazzy exotic barrier option valuator rather than the more accurate t distribution because the Gaussian keeps telling you you're making money but the t says hey, you're deep in the grey there, pal, and you're gonna get bit if you keep making this bet!

The math works (most of the time) if you use it correctly. What we need is to get back to a point where massive failures really can be blamed on pure chance or a bona fide failure of the math, instead of willful ignorance in the face of unbearable temptation.

[Just to be clear...I am NOT defending UBS in any way, shape, or form...but I'm also not condemning them. What happened is endemic to the banking industry as a whole, not UBS specifically.]

Re:You think?

[NoOneInParticular]

So please tell me, what is the financial sector using these days, if not risk models that are ultimately dependent on a (log-) normal assumption? What fundamental shift has happened that threw away 50 years of economic theory forming? What has replaced it? And as for smart people missing the bleeding obvious: Mandelbrot showed empirically that they were completely wrong more than 40 years ago. This was highly publicized, accepted as correct, and the smart people ignored this because Gaussian fairy land is a very convenient mathematical framework to publish in. I would be very surprised if the quants these days work on fundamentally different assumptions from the academics.

100% accountability? I'm not that's what they want

[SebZero]

I think that such a rigid system would prove to be a double-edged sword and would ultimately not be adopted in some institutions. It requires precise tracking of what each trader "bets" vs their losses and the application of rules to stop them from losing too much. This is data that can find its way into the outside world in the case of scandals such as this and I'm not sure investment banks would want a perfectly documented account of losses becoming public. They play a game of high-stakes risk on a daily basis, under the respectable cover of expensive premises and thick financial service books.

A friend of mine who is in what most of us would call an extremely well-paying profession told me about a highschool friend of hers who worked as a trader in London and retired at the ripe old age of 42 to live in an amazing appartment with waterfront views in central London. I was grumbling about banker-types making phenomenal money and being nowhere near as intelligent as doctors/lawyers/engineers, to which she replied, "Of course they're not the brightest, they're certainly not dumb, but they're wired different to you or I - they're risk-takers. What they do with large sums of money on a daily basis, is gambling in a casino where 'the house always wins' isn't always the case. Normal people put in their position would not take the risks they take for fear of losing"

You can potentially win big, lose or stay the same - but some of these institutions trade retirement funds or government health funds. Governments tend to have inquiries if things go wrong and not having an exacty record of how a system broke down allows the bank face while they use the trader as the scapegoat for everything that went wrong.

No on can catch them when it matters

[RobinEggs]

I remember a highly compelling thought experiment published in the New York Times editorial page, back when the economy was first going south - a though experiment which, the authors later revealed, was more real-world practice than experiment.

The way big risks and 'trendy' trades often work is this: if a certain method of investing your pool looks good and other traders in the firm are making big money on it, you have two choices. You can go against the prevailing wisdom, but should you lose then you're the idiot who lost $10 million and you're fired. You can also go with the trend, in which case either you all win together or the entire company (or entire economy) goes down in flames and you're no worse off - in reputation or employment - than anybody else.

My point for bringing this up is that it's not about random idiots going nuts; even if that happens it's just one idiot or one company that dies, and at least they're all the 'rogue' idiots are effectively sociopathic, conflicting entities. The real damage comes when every trader agrees on things; sooner or later it's all coming down, and the rest of us go down with them.

So frankly I think it's *great* news that this guy lost $2 billion; at least it means UBS isn't so locked-down that individual traders can't take risks. Better they learn from giving traders too much freedom than we all learn from them being given or being taught too little.

Re:No on can catch them when it matters

[garyebickford]

It was not good for UBS, but IMHO it was good for the system. This is an example of moral hazard in action - that principle that was violated by at least some of the US bank bailouts.

As for the size of the loss, forex trading tends to be very highly leveraged, so it's possible to have a hypothetical risk of 100 or more times the amount of a transaction - but with a very low probability of that risk occurring. The risk is not a single number but a probability density function (like a gaussian) with a very long tail (where the 'black swans' live). I presume that the bank's risk analysis system was not correctly recognizing the potential risk and requiring the appropriate hedge .

I certainly don't know the specifics of this incident, but black swan events are very difficult to quantify and it's a delicate balance between being overprotective (leaving money on the table) and being underprotective (leaving the bank vulnerable). A very small difference in the 'number' chosen can make a huge difference in the result. And to make things worse, that assumes that one has thought of and correctly judged the impact fanout of all the possible relevant risk factors - should an asteroid impact, or a large earthquake under London be included in the computation?

Re:No on can catch them when it matters

[teknomad]

At a fundamental level, a bank is nothing more than an engine of greed. I don't mean that in a derogatory sense: I mean that a bank is a simulation of the human emotion of greed. As such, it has the capacity for both great good and great evil. Left to its own devices, a bank will naturally seek to absorb all available value in an economy for its shareholders -- that's what I mean by "a bank is an engine of greed". Because of that, banks critically depend on regulations to tailor and guide their behavior to optimize their positive effect on society (banks are really good at another human trait: following rules). A bank's natural tendency will always be to push against the regulations that society places around it, so the balance between regulation and greed is where society needs to focus to measure whether a bank is doing good, or needs more regulation,...or needs to be disbanded.

Traders have a nasty job...they have to ride that thin line between regulation and greedy actions. If they step too far one way, they're "rogue" and need to be put down like a rabid dog...if they step too far the other way, they're wimpy and ineffective and need to be fired. The problem with riding that thin path is that it has lots of hidden potholes and stumbling blocks that can easily pitch the trader to one side or the other. So a big part of a trader's skill is to recognize when they've stumbled from the path, how far they can stray, when to stray, and how to get back on the path.

If I think about a few of the past rogue traders (Nick Leeson, the LTCM crew, Kerviel), the one thing they all had in common was that they started out with their hearts in the right place: they all strayed outside the thin path between regulated investment and unregulated greed in order to benefit their firms but they were unable to adjust their course in time before things got out of control. HOW they got there is what should really concern us, because it points us to the types of new regulations that may help tighten up the line and smooth out the path so this happens less often with less impact. In a sense, then, I agree: UBS's rogue trader is an indication of evolution, not a singular damning event.

What happens next will determine whether we as a society learned anything from UBS's failure, or whether we are content to keep evolving along this same line.

Re:No on can catch them when it matters

[TapeCutter]

where the 'black swans' live

Same place as me? - All swans are black around here.

why even let them exist

[Spy Handler]

the world was doing just fine without these "investment banks" gambling with billions of your money while dreaming up drivel like "mortgage-backed security derivatives"

I'd shut them all down and put the CEO and CFO and the rest of the O's, all of them, in prison.

Blame the bank administration

[msobkow]

The blame should be placed squarely on the shoulder's of the bank administrators. There is absolutely NO EXCUSE for not noticing a 2 BILLION DOLLAR LOSS.

It's not the computers.

It's not the traders.

It's not the system.

It's the BANKS ADMINISTRATION.

And it's high time that those lazy incompetent greedy bastards were held RESPONSIBLE for their incompetence instead of getting "bonuses."

Re:Blame the bank administration

[MartinSchou]

And it's high time that those lazy incompetent greedy bastards were held RESPONSIBLE for their incompetence

They will. Oh, how they will! This year they'll only get 95% of the bonuses they got last year. That'll teach them!

a hedge worth having

[epine]

The obvious thing is that you are supposed to be hedged.

You need to imbibe some Argumentative Theory, followed by a Black Swan shooter.

There's a mathematical definition of hedge, and there's the social theory of hedge. The later means "but I think I can get away with it, so it's OK". The mathematical version depends on having correct variance models. If you don't, no hedge exists. Taleb 101.

Society would benefit from hedging itself against the tendency of bankers to hedge themselves deep into the grey zone.

Seriously, bankers talking about risk is a lot like Tom Cruise interviewed after filming Days of Thunder appearing to say--very fervently--that the idea from the movie that you can't control circumstance at 200 mph is full of baloney and that he really got mad filming those scenes where other characters throw this in his face. He races his own cars and believes in control over destiny, which is common among people who take insane risks.

Even if you have LTCM wonks dictating algorithms to be coded by nuclear power engineers and run in NSA bunkers, you can't escape precipice risk. But you can shepherd all the risk with your border collie safeguard systems into the universal millisecond of doom.

So in the lingo, an unhedged risk is the one where only one bank has egg on its face, and a hedged risk is where every fucking bank has egg on its face, and greater society picks up the tab.

Re:a hedge worth having

[WoOS]

So in the lingo, an unhedged risk is the one where only one bank has egg on its face, and a hedged risk is where every fucking bank has egg on its face, and greater society picks up the tab.

This is the general problem with short-term speculation. It is a (near) zero-sum game (the near due to possible arbitrage gains from providing liquidity for long-term trades). Yet for years banks have posted record gains out of it. That money had to come from somewhere. It came out of the coffers of the financial institutes which had to be saved by governments in the last financial crisis.

And had UBS lost not only 2 but say 20 billions, it would too have to be saved (again) by the Swiss government. Yet the 20 billions would have turned up on several people's bonus/dividend slip.

Don't believe what the official report is

[roman_mir]

The original report is most likely nonsense.

I don't believe that this was just what is reported.

Yes

[Kupfernigk]

I'm sorry, but all the smart mathematicians and physicists in the world can only derive better models and better algorithms. What the financial sector seems to have failed to do is to employ enough psychologists. Taleb isn't wrong; he has pointed out that the models cannot account for random political events, for psychological forcing, or natural disasters. The probability of these cannot be reliably assessed.

On the other hand, psychology has a lot to say about the arrogance of psychopaths who assume that everybody else is less clever than they are, and how this is usually the factor that leads to their downfall.