Join Mailing List

For latest news and information about Treasury and Financial Markets, enter your details below:

Discussion Paper 10/4

Print Preview Send to a Friend Share

Published: 3rd September 2010 by William Webster

This DP raises questions about the capital charge for trading books. Look more closely and you will find that the regulation of investment banking by the FSA is ineffective. Furthermore there is no reason to think that in the future it will improve.

Investment banks will continue to run risks that periodically threaten their solvency. Their current interconnection with the retail deposit business means government support is guaranteed.

A pragmatic solution would be to disentangle investment banking from the retail business.

This free market approach is much favoured by investment bankers for other industries. Good business flourishes and bad ones die without placing a burden on the taxpayer.

Now let's see why regulation won't protect us.

1. Regulation does more harm than good

Why do trading assets make up a large percentage of banks' total assets but at the same time account for a much smaller percentage of capital?

Because in the short run markets can only fall by so much and therefore we can control trading losses before they get too big by closing out positions.

It means that the amount of capital held against trading positions need not be too great and capital can be leveraged to a greater degree - good news for profits and bonuses.

It's a trader's winning formula. Gain as the market rises and use stops to protect the downside risk. The only problem is it doesn’t work. What happens when markets fall?

You become a long term investor and you face greater losses because the holding period is extended. So how much more capital should banks' hold for trading?

On average, changes made by the Basel Committee on Banking Supervision (BCBS) will increase the capital held against trading activities in large banks to more than three times current levels. (DP10/4 s. 1.1)

But you have to ask yourself can regulators ever accurately determine the future capital a trader needs?  It means being able to predict with a good degree of certainty how much can be lost.

How do you do this if stop losses don’t always protect the trader because the market gaps down?

This risk is incidentally increased by regulation. Why? Because only the biggest firms are considered to have the appropriate risk infrastructure to trade. This leads to market concentration. Where can a bank with 20% market share offload its positions without making the position a lot worse?

I think we know the answer to that. It's to you as the taxpayer.Let's avoid regulation that causes structural changes and increases liquidity risks.

Why not let trading flourish in firms that are suited to it - those without recourse to taxpayers?

2. Can regulation see what's going on?

Back in the 1980's some banks extolled the virtues of derivatives because profitable trades went to P&L and losses were investments.

25 years on the FSA reports:

"26% of all losses, by value, were in positions that switched from trading book to banking book ........ but there was no evidence of loss-making positions moving from the banking book to the trading book during the period".

And then suggests many firms operate "material investment banking operations in the banking book". (DP 10/4 s.5.8)

This suggests that banks are trying to reduce capital charges for market risk. Does it matter? It does if they are too big to fail because you end up paying out and that's exactly what's happened.

Of course regulators will never know all of what's going on in complex businesses so in reality the amount of capital that's truly required is always going to be an estimate. 

This uncertainty means complexity needs to be kept out of retail deposit taking.

3. "Prohibited" a word regulators need to apply

There has long been a principal in markets that traders don’t value positions. It's called segregation of duties and it's there to prevent fraud. Small firms are regularly hauled up for weak controls in this area but what does the FSA see in larger banks?

  • Opaque valuation methodologies
  • Failure to value illiquid positions
  • Misapplication of  bid-offer adjustments

Despite these basic problems the FSA proposes new capital requirements based on:

Valuation uncertainties - where the value of cash flows on a risk adjusted basis varies because of methodological uncertainties and

Supply/demand uncertainties - where changes in market liquidity can lead to serious discrepancies between reported valuations and those based on actual transactions. (DP10/4 s.6.7/6.8)

We've known for a long time that the valuation of cash and derivative positions has been inconsistent and some cases creative. Why has the regulator failed to clamp down on these shortcomings rather than promote an academic debate?

Take a look what the Bank of England says on CDO valuations:

"In April 2008, the Bank of England reported that the average valuations of six large financial institutions’ super-senior tranches (rated AAA at inception) at the end of 2007 ranged between approximately 80% of their notional amount to as low as almost 20%. This indicated losses of between 20% and 80% of notional. The capital required to be held against floating rate AAA notes such as these was around 1.6% of notional". (DP 10/4 Box 5.1)

The solution is simple.

The valuation uncertainties that apply to complex securities should be removed from banks that have access to retail funds by using the word "prohibited".

4. No to VaR

Regulators allowed firms to use their internal VaR models to measure their regulatory capital requirements.  With hindsight we know VaR doesn’t work in stressed conditions and it's why in 2009 changes were made that increased the capital requirement of trading books using stressed VaR.

A robust regulatory approach would completely override VaR. It would ask how much can the trading book lose every 20-30 years based on illiquid markets and a 3 to 5 year holding period without correlation benefits. It's a stress test that makes a trader an investor and it's valid given our 2007-8 experience.

Of course it's obvious that under these conditions the capital requirement of trading books (even those that we previously regarded as being low risk) would make trading uneconomic. That's a problem.

If the regulator did its job then banks continually hold grossly inflated capital against trading risks ready to occasional absorb extraordinary losses. This would substantially reduce returns but without this buffer the taxpayer will always be exposed to "tail risks" and the occasional need to foot the bill. But it need not stop trading.

It just tells you that trading should be separated from retail banking activities. It should take place in separately capitalized firms that explicitly have no recourse to their sponsors or any public guarantee.

5. Credit trading? Not with my deposit please.

84% of credit derivatives are between financial institutions. (DP10/4 s.4.26)

The game of pass the parcel is between banks and when everyone goes for the exit at once the market is illiquid.

This doesn't matter provided the credit trading business is relatively small. But that's not where we are. It's a concentrated market where investment banks can lose a lot and you may end up with the bill.

The regulators that were convinced that credit trading was a "good thing" because it dissipated risk now want higher capital charges. Who knows whether they will be accurate? After all just how do you measure correlation risk?

Credit trading isn’t bad it just needs to be kept well away from your deposit money.

6. The final irony

The FSA accepts that where banks can use internal models then diversification and offset is permitted whereas under standard models offset (which lowers the capital charge) is limited.

So what happens when the FSA discovers that a bank has risk management deficiencies? In theory the regulator prohibits the use of offset and this increases the capital charge and makes trading more expensive.

But the FSA acknowledges "In extreme cases, the consequence of removing a firm’s permission to use the model could threaten the solvency of the firm, which can limit the set of tools realistically available to regulators when deficiencies are found. This problem is compounded as model deficiencies tend to be revealed in stressed times when further capital-raising would be difficult". (DP10/4 s.3.27)

In other words "we can't do anything when a bank is too big to fail". It's fundamentally what's wrong with the current regulatory regime.

Displaying 1 to 5 of 5 results in total.

Related Documents

Registration RequiredConsultation Paper 12 31 100% relevant

11th January 2013

Removing the simplified ILAS BIPRU firm automatic scalar increase and other minor changes to BIPRU 12. Although a consultation paper this will provide some comfort to simplified firms. The relevant parts are:


Free to ViewFTP Discussion Group 83% relevant

7th February 2012

The presentation for the discussion group held at Mutual One on 25/1/12 and 3/2/12.


Registration RequiredRegulation > Consultation Paper 09/17: A Specialist Sourcebook for Building Societies June 2009 62% relevant

1st August 2009

The general thrust of the CP is that Societies must prove that they have both the management and the systems capable of effectively dealing with the risks they face. This is part of the enhanced supervisory approach now adopted by the FSA. It states that systems and controls must match the level of complexity in a firm's business model. The FSA will adopt a more interventionist approach in order to ensure this is the case. The proposal is that building societies and the regulator will determine whether the risk management policies adopted are appropriate. Where they are not the Society can either simplify its business or improve its risk management. The FSA also intends to limit societies diversifying their business without a full assessment of capital adequacy. The FSA has considered applying similar CP 17 guidance to the banking sector but has decided on account of the "lack of homogeneity" that this would not be practical and in their case a firm-by-firm approach is more appropriate. The CP addresses treasury and lending. It contains five approaches to treasury management, three areas of treasury guidance and three approaches to lending. Consultation closes on 5th September 2009 with implementation due in early 2010 when a new Building Societies Sourcebook (BSOCS) will replace IPRU-BSOC.


Registration RequiredRegulation > Consultation Paper 08/22 Strengthening liquidity standards December 2008 61% relevant

31st January 2009

This CP sets out the FSA's plans to reform the liquidity regime. It requires firms to undertake a much more rigorous analysis of their liquidity position. This includes the effect of stressed conditions on their business. The firm will submit what it considers to be an appropriate liquidity buffer to the regulator. The FSA will then decide whether it is sufficient. In determining the buffer the FSA will also assess the firm's systems and management. If these are considered weak the buffer will be increased accordingly. The liquidity buffer can only be held in liquid assets. The FSA's view is that this primarily means Gilts, sovereign debt or central bank deposits. The FSA makes it clear, "The responsibility of adopting a sound approach to liquidity risk management is on firms and their senior management".


Registration RequiredRegulation > Consultation Paper 08/24 Stress and scenario testing December 2008 61% relevant

12th April 2009

The FSA wants to see more stress and scenario testing in firms. Senior management should be involved. Previous assumptions have been too relaxed. Stress testing should be in detail with the mitigating actions rehearsed. Reverse stress testing is introduced as a method of identifying critical events. The FSA is not going to tell you how to do this, it's up to you. However firms can expect greater challenge on the assumptions made.