William T. Ziemba

graphics/ziemba.gif

Professor Ziemba is Alumni Professor of Financial Modeling and Stochastic Optimization at the University of British Columbia. He has written and co-written many books on financial theory and practice, of which the four below are the most recent.

Books

Security Market Imperfections in World Wide Equity Markets , CUP, 2000

Worldwide Asset and Liability Modeling , CUP, 1998

Stochastic Programming: State of the Art , Baltzer Science, 1998

Finance , North Holland Handbook Series, 1995

The Efficiency of Racetrack Betting Markets , Academic Press, 1994

Lessons from the theory of gambling

Only bet when you have an edge.

You can never win betting situations with no advantage. Doubling up, or martingale systems, while highly touted, are flawed and will not work. You will have many small wins then a large loss that will wipe out all your gains and more.
Those who win on every trade are losers or liars.

No strategy will provide profits in all scenarios except true arbitrages, which are hard to find. The aim is to generate consistent profits over time and that will involve taking some losses. Successful traders take small losses on losing trades and make large profits on successful trades.
Know what you expect to gain from a trade before you enter.

Predetermine your exit point before you get into a trade - and revise this as new information unfolds. Discipline and organization is the key to successful implementation of successful trading strategies.
Do your research.

Fully research your strategy both in theory and with data, and update this constantly. Financial markets are partially predictable. Models that attempt to predict future asset prices need updated data constantly and model refinements periodically.
Truly diversify.

Correlation matrices are scenario dependent. Simulations around historical situations do not measure the true correlations . For example, bonds and stocks are generally positively correlated, but in crash situations they are negatively correlated.
Do not over bet.

Determine in advance the bad scenarios that could occur and how much you will lose in each case. Remember that the true size of the current bet, which is constantly changing, depends on the hedge ratio, namely, the change in portfolio value that occurs because of a change in the underlying index. What was a small futures derivatives position and fully risk controlled can turn quickly into a large position that amounts to overbetting. Hence the risk position of the portfolio is constantly changing and must be controlled carefully .
Evaluate the impact of all scenarios.

Evaluate all scenarios that could possibly occur, even those that have not occurred yet. Use similar situations to determine reasonable scenarios. Know in advance what to do if a particular scenario does occur.
Have deep pockets.

Have enough capital to withstand crises and use that capital to avoid these crises and profit from such crises. During crises , cash is in short supply and those who have it both avoid disasters themselves and profit from being able to buy/sell securities at advantageous prices. Calculate how much capital is needed to avoid disasters in extreme scenarios given your current level of investment bets. Since the latter is changing, the size of the bets must also be constantly monitored and changed.
Accept small losses.

Try to avoid large losses. Exit markets with small losses and preserve your capital.
Follow a risk control system.

Focus on not losing; it's more important. Focusing on winning can lead to massive losses as many traders and hedge fund managers have found out.