Risky Business: A Crisis for Money Market Funds

by Ingmar Adlerberg, President, Riskdata.

For corporate treasurers, as for many investors, money market funds are the Fort Knox of investment vehicles. By purchasing debt securities issued by banks, large corporations and the government, money market funds carry a relatively low default risk while still offering high returns in comparison to similar low-risk/liquid products. However, given the current credit crunch, these funds are experiencing unprecedented volatility. Some of those funds may have been investing in riskier holdings that were not in line with investors’ expectations.

Poor understanding of risks led to losses

For example, unexpected losses by French money market funds, in the summer market crisis, caused by poor understanding of exposure to mortgage backed securities and credit, are now a major concern for corporate treasurers and CFOs. Of the total fund universe, some €500bn were invested in money market funds - the typical types of fund in which corporate cash is invested on a short-term ‘riskless’ basis. Yet August saw a major crisis for money market funds. In France, OPCVM sudden losses accounted for more than 50% of year-to-date performance. According to the most commonly used risk models, this type of event should occur once in thousands of years.

A study carried out by Riskdata, the leading provider of risk management tools for the alternative investment market, provides detailed analysis that shows that neither basic risk measures, nor fund ratings succeeded to identify the nature and magnitude of risk.

Riskdata analysed the performance of 2,000 funds on the French market with total assets under management of €1.368 trillion (thus covering most of the volume of the French funds market, which is €1.6 trillion according to official data). The funds studied were grouped into categories based on their stated strategy from completely safe to more aggressive (using the categories of the major fund rating agencies), and Riskdata calculated performance and risk for each category for the year-to-date and for the crisis period from July 15 to August 20.

The analysis revealed two developments which illustrate a disparity between what investors thought they were getting, and what the money market funds delivered, or in many cases didn’t:

1. Traditional risk tools used by investors were insufficient.

2. Rating agencies’ classifications of funds do not always accurately reflect risk

• Of 333 money market funds, a few suffered losses totalling €17 million during the summer crisis, although the vast majority of funds in this category remain robust and did not suffer any losses.

• In the treasury dynamic category incorporating 122 funds, 39% lost money in the crisis, with losses of €393m. The average loss in the crisis of the funds that lost money in this category was -1.15%. The year-to-date performance of those that lost money was 1.05%, so over half of their yearly performance was lost in the crisis. The average year-to-date performance of this category as a whole was 1.7%.

• Of the 88 funds in the treasury dynamic plus category, 50% lost money in the crisis, with losses of €119m. But the average year-to-date return of these funds of 2.1% was superior to the treasury dynamic category, proving that labelling funds in such categories gives no indication of the actual risks being taken.

[[[PAGE]]]

1. Traditional risk tools are necessary but insufficient

Traditional tools to guide investors to select true risk-free treasury instruments such as volatility or even Value-at-Risk gave underestimated predictions compared to actual risks. Moreover, they did not allow the differentiation of risky money market funds from classical low risk ones. The average two-week volatility of dynamic money market category running up to the crisis was 0.2% for the ‘losing’ funds, and also 0.2% for those that did not lose. The two-week Value-at-Risk (at 99% predicted loss) was -0.49% - not much different from the much safer treasury category.

This result confirms that classic risk measures are necessary but insufficient, as they do not allow investors to see the ‘colours’ or types of risk in order to carry out informed investment decisions. Also, these measures are not insightful enough in extreme crisis situations.

2. Rating agencies’ classifications of funds do not always reflect risk

Riskdata analysed the ratings of funds by a major rating agency and discovered a poor relationship between the fund rating and the real market risks being taken. In some categories of funds, funds with higher ratings were actually more likely to have experienced losses.

While risk assessment probably is not the only purpose of fund ratings, this is a concern, and rating agencies would benefit from developing ratings that better reflect risks.

Adequate risk tools could have prevented losses

The research carried out by Riskdata illustrates that with the appropriate tools, it would have been possible for investors to avoid taking unwanted risks. Riskdata studied fund performance using the data available before the crisis, and found that funds displaying high credit risk could have been easily identified. This also applies to anticipating the magnitude of their losses. Alternative investment risk management approaches provide a more objective and robust way to understand risk. [[[PAGE]]]

With the information available in the beginning of July, an investor could have taken precautions by splitting all money market funds into two categories based on risk as predicted by Riskdata: the first containing significantly higher credit risk - which mirrored the funds that actually proceeded to suffer losses, and the other - with minimal credit risk, containing funds that emerged from the turmoil relatively unscathed.

Classic risk measures are necessary but insufficient.

The orange line on chart 1 shows the history of a dynamic money market fund. The blue line is the credit factor that was discovered as the prevailing risk factor, using Riskdata’s risk profiling technique. The chart illustrates that this is what effectively drove the loss when the credit factor in question (low grade credit) jumped. The light grey shading shows the period of nine months preceding the July crisis covered for the analysis.

The spider chart on the right shows Riskdata’s estimated risk profile of this money market fund. The figures in the table on the left show the real and estimated losses for four different money market funds.

Conclusion

Alternative investments have highly demanding risk control requirements and imperceptible risks cannot be identified using simplistic approaches. Just as someone’s health cannot be determined solely by his or her body temperature, risk cannot be reduced to a single number. The crisis highlights the need for better understanding of risk ‘colours’ and the type and real amplitude of risk undertaken by investors and treasurers.

While there will be calls for greater regulation, there should also be calls for better risk control practices and tools. Financial market innovation will always outrun classical regulation. Everybody in the market - corporate treasurers, investors, regulators, asset managers - would strongly benefit from the best practices and most advanced risk management tools. The risk framework developed for alternative investments needs to be more broadly adopted, in terms of systems, models, process and regulation.