Updates

Asiahedge 10-2005

ASIAHEDGE OCTOBER 2005 ISSUE

INTERVIEW

QUANT ASSET MANAGEMENT (QAM)

Fear and greed are not the most effective tools for successful hedge fund management, but all funds managed by human beings are to some extent influenced by such strong emotions. Although many hedge funds use quantitative analysis and quant models to help run a portfolio, few free themselves completely from the influence of human involvement in the way that Quant Asset Management does.

Singapore-based QAM's Asian Equities and Global Equities are both run on a pure quant basis using dynamic models based on a blend of 20 to 30 factors with the underlying conviction that share prices are driven by earnings expectations and relative valuations and that markets are semi-efficient.

The funds were launched in April 2004 and as of the end of September they are both up approximately 40% since inception. In September, the Asian and Global funds clocked up an estimated monthly performance of 8.72% and 10.64% respectively.

If there was a key word for this fund it would be 'dynamic'. " The dynamic weighting process of factors is the key to our fund," says Frank Holle, who together with Chatchai Ngampakdeepanich set up QAM in 2004. "The reason we can run the fund with no human interference is because we use multi-factor models that are dynamic. On a frequent basis the computer will look at what's happening in the markets, then the portfolio is rebalanced on a monthly basis and the factors get re-weighted," he says.

The word quant usually conjures up images of a black box, but QAM reveals as much if not more about the drivers behind the fund's performance than a more conventional long/short equity fund.

"Our multi-factor model is looking for value stocks with earnings momentum (growth at a reasonable price). Sometimes the market directs us more towards growth, sometimes more towards value but we always work with fundamental factors related to earnings and valuation, we do not use technical analysis. The algorithms behind those factors are secret but we are pretty open about the factors themselves and can show investors historic portfolios. Every month we publish the top three sectors and countries and the aggregate momentum and valuation weightings," says Holle.

There are three criteria for a factor to enter the model- it has to make commonsense; it has to work for a big group of stocks; and it has to be reliable over long time horizon and stand up to 11 years of back-testing. Since the funds were launched, just one new factor has been added, which coupled with the funds' strong performance, indicates that the original model was already sound. Chatchai, who was with financial database giant Thomson Financial from 1996 to 2004, is the brains behind the computer program.

The models are divided into four main groups- momentum, valuation, signal combination and stock types. Core momentum factors are analyst revisions, changes in consensus earnings, price changes and dispersion in earnings forecast. Core valuation factors include earnings growth, price to book value and dividend yield. There are 6,000 stocks in the global fund's universe and 1,700 in the Asian fund's, selected on the basis of market liquidity and analyst coverage. One of the databases used by the quant program to track these stocks is Thomson Financials which collates the views of 50,000 analysts worldwide on a daily basis.

Every month the funds are either 0% or 50% net exposed, with zero net exposure 65% of the time, based on a dynamic hedge indicator. The funds do not short individual stocks, but short index futures instead. "We invest in 52 countries worldwide making no distinction between developed and emerging, big cap and small cap so we don't short individual stocks as it's very difficult to be consistent with shorting in this way," explains Holle. Positions are generally kept for four weeks, the point which strikes a balance between maximum efficiency for the strategy and controlling transaction costs, and from month to month there is typically 70% turnover.

One of the advantages of a quant model is the ability to perform very accurate back-testing and the fund's models have 11 years of back-tested data behind them, a useful tool for calculating risk as much as it can calculate expected returns. Holle makes no apology for the strategy’s risk profile with targeted volatility of 20% and 15% maximum drawdowns. "Within that risk framework the investor can expect exceptionally high returns- the target for both funds is a 40% average annual return. The actual monthly returns and volatility entirely match the back-test results which means the back test results are reliable. When we showed everybody the back testing results one a half years ago I was quite surprised by the tame response but now we also have one a half years' real-time performance and investors are very interested," he says.

With the models already in place, QAM's goal is now to focus on researching more factors and on faultless execution. As operations and back office functions are also highly automated, QAM's operation is highly scalable, says Holle. A European fund will be added next. "Because the method is exactly the same it doesn't require any more personnel. We're now $40 million in assets under management but at $500 million the processes are the same," he says.

The Asian fund currently has AUM of $20 million and will close at $100 million. With 40% leverage and anticipated returns of 40% a year, that will enable the fund to organically reach capacity two years later at $300 million, about the maximum that liquidity in Asia will allow. "Capacity for the Europe fund will probably be about $200 million, but the global fund can be big- up to $800 million. That's ambitious, I know, but we will grow the business step by step," says Holle.

« BACK TO MORE NEWS