王章佑 - 財經

Short-Term Asian Trader

Vegasoul Capital in Hong Kong has generated notable returns for investors during the two years it has been trading its high-frequency futures program.

"We're a CTA located in Asia, trading global markets. There aren't many CTAs in the regions doing that successfully."

Meet Vincent Wong as he explains why his Vegasoul Fund has an edge over its competitors in the Asian market. "We come to the game with a different perspective than other CTAs located in the U.S. or Europe. We probably have more Asian market exposure than any CTAs out there, including some niche and emerging markets that are not accessible to them because of regulatory restrictions, such as the local Taiwanese stock index. Our first systems were initially developed on Asian futures data, and it still
surprises me that the same systems seem to work across global markets."

In the two years it has been accepting outside allocations, Vegasoul has amassed $70 million under management, equally sourced from European and Asian institutions, funds of funds, banks and high-net-worth investors, and has generated a compound annualized return of 49%.

Mr. Wong opened Vegasoul Capital in 2004 in order to provide an infrastructure for his personal trading. He had returned to his native Hong Kong three years earlier after graduating from Monash University in Melbourne with a degree in economics, and with money borrowed from his family had begun experimenting with the futures trading program he'd developed. The trading, which soon expanded to 40 markets, went well, and after three years was taking up most of his waking hours.

"Being caught up with the routine maintenance of the systems," he says. "I had no time to focus on the most important aspect of trading, which is research. Vegasoul was launched not so much to manage investors' money, but to provide a platform to expand the scope of my personal trading."

Vegasoul invested heavily in IT and building a top-notch research team, which now includes several Ph.Ds. When the process was finished, Mr. Wong decided to consolidate his accounts and raise outside capital for the Vegasoul Fund, which started trading in June 2005. Most of his own capital is also invested in the fund.

Asset Allocation Is Key

Vegasoul is predominantly a high-frequency trading program. Its 100% systematic program has two main components: 80% to 90% of assets are allocated to a short-term component with an average holding period of less than one day, and the rest goes to a long-term trend-following module.

The short-term component, which has both trend-following and mean-reversion elements, tries to capture short-term momentum that manifests itself as a trend day: the market opening on a high and closing on a low, or vice versa. Although easy to spot on a chart, trend days don't occur often, says Mr. Wong. The problem for the trader is to keep the losses as small as possible while waiting for those days to materialize. "We are old-school in this sense," he says. Vegasoul uses pattern recognition and
trend-following futures to identify the probability of such days happening.

For the reversal part of the short-term component, the firm's traders try to find times when the trend days don't materialize, when false breakouts may occur. These abortive trades will have a very good risk/reward. So if the trend-following trade doesn't work out, the reversal usually does; together these two elements have a slightly negative correlation, nicely complementing each other at the portfolio level.

The long-term trend-following component of Vegasoul's program, which is not correlated with short-term component, is unique in the sense that the strategy is embedded in the portfolio construction process, says Mr. Wong. "We have found that entry and exit points are relatively unimportant, and any reasonable trend-following strategy will provide you with an edge. By way of analogy, it's close to impossible to build a better mousetrap. Instead of developing a standalone trend-following model like
everybody does, we have incorporated it as one of the factors that influence our asset allocation. Hence, our asset allocation algorithm is the strategy itself."

Mr. Wong says that allocating capital across markets is the most critical investment decision at Vegasoul. "We have developed a robust asset allocation algorithm to do this. The traditional problem in mean-variance analysis is that portfolio weights are highly sensitive to expected return estimates and hence can be extremely unstable over time, limiting this cornerstone of portfolio theory's applicability for practitioners."

He says there have been many attempts to remedy this problem in academic literature, but Vegasoul has found them to be unsatisfactory. "We believe we have found an elegant solution that takes into account all the factors that a practitioner would face in the real market. The process mimics evolution as it happens in nature, and the portfolio is self-adaptive to changing market environment. Changes to the weights are size increments that are relative to the statistical significance of recent data."

The fund is active in 80 or so markets, covering a broad range of asset classes, including currencies, fixed income securities, equity index futures and agricultural-related securities and derivatives. It does not cherry-pick markets, but will trade any one that has sufficient liquidity.

Toward Multiple Strategies

Mr. Wong says he is not actively raising assets at present. "We should put our focus on building the track record, and if we have the results, the money will just come in. Marketing, per se, will not increase the fund's return, and is probably negatively correlated to it."

As to the future, he says he sees a lot of interesting developments on the horizon. Vegasoul is actively developing an equity-trading model and an option-trading model. The firm's goal, says Mr. Wong, is to break out of the confines of a CTA box, and develop into a multi-strategy fund.

Vegasoul Capital is staffed by 23 people—Mr. Wong considers the firm's size another one of its distinguishing characteristics. These include seven people in trading; two in back office; four in research and the rest in IT. The firm's head of research, Lewis Chan, has a doctorate in economics from Harvard University, and senior research analyst Andy Leung has a doctorate in physics from Cambridge University. Mr. Wong is involved in both the business and the research aspects of the operation.



--