Before starting to trade, we need to have clear goals. The naive answer is “to make as much money as possible”, but this doesn’t help us answer the most important questions:

- how much money will we make?
- how much money do we need?
- how much money are we willing to lose?
- when is a strategy “good enough”?

Without knowning how much money we will make, there is no way to compare our trading strategy to alternative ways of investing our money. There is also no way to determine if it is worth putting any time into researching and developing algorithms.

Without knowing how much money we initially need, we are unable to to make sure our account is properly funded. This question ties back to the one before: typically the money we make is somehow tied to the money we have available for investing.

Without knowing how much money we are willing to lose, we will be unable to decide if a strategy is acceptable to us; in particular if we are financially and mentally prepared to execute this strategy. Note that the amound of money we are willing to lose is typically smaller than the amount of money we need to execute a strategy.

Without knowing our criteria of “good enough” we have no way of determining when we are ready and prepared to go live with our trading strategy. There is typically no black & white answer here. Therefore, this question is tightly linked to the question of how to compare and rank strategies.

It took me a while to come to a good understanding how to answer these questions. These answers reflect my personal preferences and your answers might be quite different than mine. Here are my targets:

- average annual return of about 20% of the total account size
- maximum draw-down of less than 25% of the total account size
- smooth equity curve, without dips or sudden jumps, maximum flat period of 6 months
- diversified investment
- backtested results spanning more than 5 years
- proven correlation between simulated results and realtime trading spanning more than 1 year

To illustrate this a little bit, see this equity curve from one of my algorithms:

The algorithm automatically picks a handful of stocks from the S&P-100 and Nasdaq-100 for a diversified portfolio. The average annual return is about 17% over the past 10 years, including $0.03 per share round-trip commission. The worst drawdown was about 20% in 2008. This drawdown lasted for about year, which is a little longer than what we would want. The equity curve looks very pleasing and smooth. The algorithm was backtested over a period of 10 years, based on 5-minute data. Overall, this is a good start.