= Subscriber content; or subscribe now to access all American Banker content.

Fraud by Rogue Traders Is Absurdly Easy to Stamp Out

There's an unending parade of proposed systems to prevent the kinds of bank fraud that attack customer accounts—every size and shape, from new companies with another way to cut the data, all the way to the ultra-expensive EMV

But when you read about a big bank brought low by fraud—such as UBS or Societe Generale—it's not because of the kind of fraud these up-to-the-minute solutions target, i.e. fraud in customer accounts.  The big damage, enough to threaten to wipe out a large bank, is from fraud in proprietary accounts.

Kweku Adoboli, former UBS trader, is now on trial for a $2.3 billion fraud, in London. "Rogue traders" used to be spectacular—remember Nick Leeson, whose fraud bankrupted Barings Bank in the 1990's? Now it's commonplace. No one's interested in Adoboli. 

Yet this is grotesque, bizarre. It doesn't happen to Microsoft or Walmart, it doesn't happen in 5100 Chase branches. Multibillion dollar frauds (like multibillion dollar risk management failures without fraud) typically stem from proprietary trading, particularly in derivatives. Nobody's proposing preventatives for this kind of fraud. That would be too simple, too untechnological, too humdrum.

Sure, it's possible for a trader such as Chase's "London whale" to cause billions in damage even without fraud, because of inadequate risk management. But trader fraud should be much easier to eliminate than losses from misguided risk judgment about transactions that are promptly, clearly and accurately accounted for. 

Improved process can snuff it.  Trader fraud occurs only because we do not require immediate trade confirmation posted infallibly to the institution's books, plus marking of positions that is uninfluenced by traders. Instead, the trader is allowed to wave a magic wand to create false transactions and false values and hide real ones.

Why don't we see multibillion dollar trader fraud on the NYSE, or NASDAQ, or bond markets? It mostly depends on derivatives, those miraculous engines of "risk reduction." It is facilitated by resurrecting antique, late 19th century methods of trading and settlement: the telephone, maybe a keyboard, and no real-time verification, no clearing. This flight back to the distant past is necessary to avoid transparency and exposure of orders to competing market makers—hence to maintain unlimited spreads. To participants in this narrowly self-serving system, the fraud risk is a small price to pay for big bonuses, especially since fraud "can't happen here," and counterparties assume they are protected by collateral.

This would be impossible without non-regulation, enacted into law in the waning days of the Clinton administration, over the incisive and farsighted objections of Brooksley Born, who then chaired the CFTC. "Greenspan didn't believe that fraud" prevention "was something that needed to be enforced."

Fraud also couldn't happen without government kowtowing to the industry's demand that we not only compete in laxity with even more supine foreign jurisdictions, but promote unregulated access by our own institutions to wild West trading arrangements, for instance in London. And it probably couldn't happen without repeated Republican slashing of the regulatory budgets.

But we're not going to get much more clearing. Nor transparency, nor reasonably efficient markets. We won't have U.S. standards applied to all trading controlled by U.S. institutions. 

We're also not going to implement the more straightforward remedy of eliminating proprietary trading by banks altogether. 

Even so, we can prevent most proprietary fraud.

Instead of having the trading records checked and values assigned to positions by a pitiful on-site flunky whose dream is to get the trader to recommend him for a low-level trading job, how about this:

Record all trader phone and data line traffic. Prohibit trading not done through these monitored lines. Have all the recordings checked within 24 hours (preferably within eight) by a remote compliance or accounting person.  The recording would identify the counterparty electronically, and the verification agent would compare each trade against whatever record the trader transmitted. All discrepancies would be immediately resolved by telephone inquiry to the counterparty. Exclude traders from the position valuation process.  Trust, but verify.

Verification and valuations agent would be shifted from day to day between at least 10 traders and would be prohibited from telephone communication with them.  Automatically fire any trader with repeated discrepancies. 

Otherwise, what happens when, two or 10 fraudsters get together. They may all work for the same institution, or, worse, they could be working for different institutions. The resultant losses could total $10 billion – or $100 billion.

A LIBOR-size crime that directly hits banks' books. If this isn't systemic risk, what is?

Andrew Kahr is a principal in Credit Builders LLC, a financial product development company, and was the founding chief executive of First Deposit, later known as Providian.


(3) Comments



Comments (3)
Right on, as far it it goes. Please note that these trillions in speculative activity wouldn't be occurring at all if they weren't government-guaranteed (de jure now). We still have effectively un-regulated and government-guaranteed speculation, allowing a small group to opportunistically cream off huge profits. That game just shouldn't exist. The old, pre-1998, structures for trading were not only not government-guaranteed but the real risk trading took place in Wall Street firms that had an owner-operator mentality with the partners over-seeing what they considered and what was their own money, not a government-guaranteed pool of funds for playtime for a few. In the earlier era, traders had adult supervision.
Posted by HarrisonH | Wednesday, October 03 2012 at 9:24AM ET
sorry for the follow on post.

But this article is seminally correct. Not one exchange or futures market has ever had a "risk" or position meltdown. They have huge diesel engines which run SPAN, and methods for covering variation margins etc are rock solid (like a truck diesel engine). Yet these same bulge brackets assuage with sophistry and hubris VaR correlation etc etc. till you wonder what you could ever do without a PhD in maths.

Well put: The rewards are so great for the status quo, that the occaisonal blow up doesn't really warrant change.
Posted by kiers77 | Monday, October 01 2012 at 1:26PM ET
"Improved process can snuff it." AMEN. Have you noticed the number and breadth of bulge bracket banks using (of all things) SPREADSHEETS as a tracking mechanism! First the trader maintains a spreadsheet (P&L), then this spread sheet is passed on or replicated independently by "front office" then it is melded against another spread sheet at the back office along with slightly more robust mainframe data, and THIS is the process used to "reconcile" cash balances.

Change an entry in any cell in a spread sheet, and you have the bank by the b*lls. 2012 and we are still talking spreadsheets here. No wonder there's fraud. Why not incorporate "notes from mom" as well into the accounting process!
Posted by kiers77 | Monday, October 01 2012 at 1:16PM ET
Add Your Comments:
Not Registered?
You must be registered to post a comment. Click here to register.
Already registered? Log in here
Please note you must now log in with your email address and password.