Priscilla Ngan Manager of Middle Office Risk Support/ Customer Indications Pricing team Operations Department Morgan Stanley Hong Kong Oct 23, 2007

Priscilla, who is working with Morgan Stanley in Hong Kong, is responsible for providing non-market risk support for Institutional Equity Derivatives division and valuation services on equity derivatives and fixed income portfolios to clients. The products covered include over-the-counter options, swaps, warrants, structured notes, exotic transactions and fixed income derivative products. The non-market risk support services include stress testing, collateral treatment analysis, event and exposure monitoring. Priscilla is the team manager of a dynamic team that works closely with risk manager, derivative sales, traders and other staff from operations. Team Latte's Rahul Bhattacharya recently caught up with Priscilla for a chat.

Team Latte :
How does it feel to be with the Operations - Middle Office Risk Support team of such a large bank? Clearly, you must be under lots of pressure on a daily basis, given such volatile markets these days.

Priscilla Ngan:
Being part of the Operations team of one of the world's leading investment bank, we need to keep up to speed and provide good quality services, investment advices and top tiered products to customers in order to remain competitive. Since I joined Morgan Stanley in April this year, I managed to maintain a work-life balanced life, by working closely with my team and resolving problems based on our own experience and the talented workforce. Although the market underwent severe turbulence in Q3 07, I was able to enjoy a life outside work.

Team Latte:
Your risk portfolio comprises financial derivatives and you need to analyze the market risks of these derivatives. How does risk analysis of a portfolio of derivatives differ from that of a cash (with linear payoff) portfolio?

Priscilla Ngan :
A cash portfolio offers more stable and less risky returns and its performance does not have high dependency on the market conditions. On the other hand, a portfolio of derivatives is more risky and the return is highly dependent on the market.

As a result of the higher risk being exposed to when managing a derivatives portfolio, a sound market risk management must be in place. Market risk could be monitored through various measures: VAR measures; back-testing; by measures of position sensitivity; and through stress testing and scenario analyses. Market risk exposures should also be managed in such a way as to maintain a portfolio that is well-diversified in the aggregate with respect to market risk factors and that reflects the firm's aggregate risk tolerance as established by the Senior Management. Further, trading positions can be monitored by employing a variety of risk mitigation strategies that include diversification of risk exposures and hedging using derivative products eg. futures, forwards, swaps and options.


Team Latte:
Before moving on to Morgan Stanley you were with KPMG, one of the big four auditing houses. Your main remit was to work on valuation of exotic derivatives and complex structured products for your clients, the financial institutions.

That work must have been very challenging, given that these days large financial institutions have asset portfolios comprising very complex derivatives and structured products? What were the main challenges that you faced?

Priscilla Ngan :
My previous role in the Financial Risk Management Division at KPMG was to perform valuations of financial products for our audit or non-audit clients. Since KPMG has the biggest financial institutions clients in HK and upon the launch of IAS 39 where all derivatives need to be fair-valued, I was given the chance to work on the pricings of various types of financial instruments, which include convertible bonds with callable, puttable and redeemable features, worst/ best-performing options with basket underlyings,etc.

Before we could work on the valuations, we had to understand how the product was being structured and the determinants of its return. In most cases, we had to decompose a single product into several individually-priced instrument and look at the return as a whole. Not only did I find the actual valuations work challenging, the most difficult part remained at times when the clients questioned our values and underlying assumptions.

As you may know, in the derivatives market, there is no unique pricing model for any product and that different models will give rise to distinct values. Further, the different underlying parameters used will also lead to distinct values. The parameters which we often found to have discrepancies include volatility, correlation, interest rates and dividend yields.


Team Latte:
A lot of financial institutions are confounded with the problem of choosing the right model for valuing a particular derivative instrument. And then there is the problem of implementation of that model, such as whether to implement the model in a numerical framework, and if so, then which particular approach to take or whether to implement it in a closed form framework.

Let me be a bit more specific: say we are talking about valuation of a callable bond with an embedded call option in the bond or valuation of a vanilla cap (call option) on an interest rate. The math model we can use to value the call option could be Black's closed form model, or we could use a Black-Derman-Toy tree to model the underlying interest rate and then value the call in a tree framework; we could also use some kind of LIBOR market model (BGM type models) and implement that math model in a Monte Carlo framework. Each of the above has its pluses and minuses, pros and cons.

In your opinion, what factors should be considered most when making a decision to choose a particular model and a particular way to implement it? The reason I am asking this is because, if an external auditor is faced with evaluating and passing an informed judgment on a financial institutions valuation methodology of derivative instruments, then how should he go about doing it?

Priscilla Ngan :
As I mentioned in the previous question, there is no unique model for any derivative instrument, it is the most difficult part of valuations.

The product you have mentioned have become a familiar investment tool for investors, thus they are easy to understand and investors are comfortable with the underlying risks (since most of the products are non principal protected).

The Black-Derman-Toy ("BDT") model is a model of the evolution of yield curve and is a one-factor model, that is, a single stochastic factor (the short rate) determines the future evolution of all interest rates. We can calibrate the parameters in the BDT model to fit the current term structure of interest rates (yield curve) as well as implied volatility derived from publicly traded options. However, the BDT model does not take into account mean reversion, an essential characteristic of the interest rate that sets it apart from other financial prices. As opposed to stock prices, interest rates cannot rise or decrease indefinitely, and as a result, interest rates move in a limited range, showing a tendency to revert to a long run value.

The Vasicek model, the first model to capture mean reversion, is another commonly used model to determine the evolution of interest rates. However, under the Vasicek model, it is theoretically possible for the interest rate to become negative, an undesirable feature. This shortcoming was fixed in the Cox-Ingersoll-Ross model, where the standard deviation factor, , ensures that the interest rate does not become negative.

The examples above were to show that every pricing model has its own limitation and so it is important that we understand the structure of the instrument we are pricing and choose the model that captures all the factors in need.

If an external auditor is to evaluate the appropriateness of the valuation methodology adopted by his client, my advice would be to first seek the client for justification of the model to ensure that the client understands the instrument he is pricing and model he used. As long as the model/ parameters used by the client are justifiable, the auditor can pass an informed opinion to accept the client's valuation methodology as reasonable.

Team Latte:
The world of a risk manager is quite different from that of a trader in any bank. Is it fair to say that the risk manager looks at the long term whereas the trader only cares about the short term? How do we reconcile these two worlds?

Priscilla Ngan:
Although risk managers and traders have different objectives, they do have to work together to ensure that aggregate market risk limits are in accordance with policies set by senior management.

Risk managers have to review the trading portfolios from a market risk perspective utilizing VAR and other quantitative and qualitative risk measures and analyses. Nevertheless, traders and risk managers also use measures such as: 1) sensitivity to changes in interest rates, prices, implied volatilities and time decay; 2) stress testing; and 3) scenario analyses, to monitor and market risk exposures.

Team Latte:
As a market risk analyst, one has to monitor adverse and extreme movements of underlying assets and then try to quantify those movements into risk estimates. How do you think extreme movements can be defined? When one has to impose a quantitative framework on defining such extreme movements - such as, maybe a three standard deviation move in the underlying asset - how relevant are they in capturing the real world risks?

Priscilla Ngan:
The global market has become increasingly volatile these days and has become more reliant on external factors. Hence, it is rather difficult to have a fixed definition for extreme movements all way long. However, say for options, volatility can be a good indication for reasonableness. The threshold for day-on-day change on volatility differs for different underlyings (whether is index or single name), length to maturity and the country/ market of the underlying. But in general, the threshold for volatility change can range from 1% to 4%.

Any comments and queries can be sent through our web-based form.

More Asia's Brightest >>

back to top