A GlobeRisk Perspective on Investment Banking
Recent years have seen massive growth in invest banks profits driven by a unique combination of factors:
- Strong stock markets
- Strong Bond Markets, especially credit instruments
- High levels of M&A activity
- Weakness in the Japanese Yen created an ideal environment in which to borrow money cheaply
Investment banks have grown and many have outstripped the capacity of their control frameworks.
Until the second half of 2007 most firms had a large structural long position in credit assets seeking to benefit further from carry and declining spreads. Whilst July 2007 has seen some unwinding of these positions, the huge inventory across the trader/hedge fund sector means that spreads are likely to widen much further. Volatility will be the order of the day with "fear" and "greed" alternating as the primary market force
A further issue for Investment banks arises from Credit Derivatives. When these were first developed in the early 1990's they were linked to specific actual traded bond instruments and volumes were small in relation to the issued credit instrument base.
As at 2007 factors are very different:
- Credit derivatives now exceed the capital market issued base by a large factor
- In addition to which, much of the volume has occurred in so called synthetic credit derivatives (instruments which trigger payment based on bond spread measures or indices)
- Modern instruments are frequently "basket" based rather than being based on a single underlying obligor.
- Securitisation and tranching (i.e. the splitting of a risk into senior, mezzanine and junior elements) have grown enormously, particularly for baskets
The net impact is that banks have valuation and risk modelling issues whilst the underlying instrument market is too small to sustain significant hedging without collapse. Arguably, this effect explains much of the widening in investment grade credit spreads in June/July 2007.
The market standard for modelling basket options is to use a single factor copula model to measure the relationship between the components of the basket. This has the advantage that tranching can easily be accommodated within the model but the disadvantage that it works badly - especially for the senior or junior tranches - must inevitably respond badly to economic cycle effects
Moving to Interest Rate derivatives, the last 10 years has seen considerable progress. The Heath Jarrow Morton model is now the most widely used full term structure interest rate derivatives model in its discrete or "LIBOR model" form. More recent innovations include the BGM model which adds in mean reversion and stochastic volatility (although GlobeRisk has long held the view that stochastic volatility models do not work well preferring "jump" style models which are capable of replicating the increase in volatility smile that normally occurs as contract expiration decreases rather then the opposite effect, as generated by stochastic volatility models). The modelling technology has enabled a large increase in volume, but only at the expense of two key issues:
- Control over mark to market valuation
- Assessing the overall risk of a derivatives book which may consist of a large position at one point in the maturity/strike/volatility surface supposedly hedged by positions at rather different maturity/strike/volatility points - GlobeRisk frequently sees situations in which the construction of the overall portfolio goes beyond the power of the model(s)
An older problem which continues to grow, despite the introduction of formal "day1 profit recognition standards" is that of "valuation adjustments", that is to say the adjustments that one need to make to inventory marked-to-market at mid price to deliver an overall valuation that is consistent with the firms accounting policy
To be fair many adjustments have been introduced since the early 1990's. However two major areas remain - in the context that an off balance sheet book essentially represent a future revenue stream conditional on certain management/IT/Operations "conditions" being met. These areas are
- An adjustment for the cost of the Tier2 and Tier3 debt required to creating capital capcity
- An adjustment to recognise the cost of the equity (and other Tier1) supporting the portfolio over its average life
For banks with long average life derivatives portfolios, these adjustment could be very large.
GlobeRisk has developed several consultancy services to assist Investment banks as they address these issue areas
GlobeRisk has unparalleled experience in assisting banks to recover from trading disaster situations, whether structural or (more commonly) rogue trader.
In 2007 we can offer the following services:
- Review of valuation controls, profit recognition and/or valauation adjustments
- Independent derivative valuations
- CAD2 program management
- Improved Risk and Performance Reporting
- Review and redesign of risk management framework and processes
- Strategy Development assistance