The Drawdown - Prashanth K

Drawdown is a statistical function that is missed by many and worse mis-interpreted by many others. Every system builder always has one eye on the CAGR and another on drawdown. It is a balance that is often unstated but has to remain rational for any system that has been developed after much effort to become operational.

A 1000 percent return with a 90-percent drawdown is a lot worse than a 100 percent return with 9 percent drawdown, although mathematically both are similar. The reason I say that the second is more important is because there is something we call: "Your worst Drawdown is yet to come", which in efffect means that no matter what percentage your system backtest would show, the chance is greater that a drawdown greater than this would arise in real trading.

Secondly, the system does not consider the psychological profile of the trader. It is tough to live with a trade where you have lost nearly 10 percent of the capital as against living in a trade where you have lost nearly all of the equity.

So, lets look at what a drawdown is and why, in my opinion, it is one of the most important figures to see in any backtest.

An Amibroker Backtest report provides two drawdown figures. They are (copy-pasted from AB Help),

Max. trade drawdown - The largest peak-to-valley decline experienced in any single trade.

As explained, it is the drawdown an open trade experiences. For example, assume your system has gone long in ABC at Rs 100 and the stock, after moving to 102, has moved to 96 without there being any additional signal.

The Trade drawdown in this case will be 6 * Qty (102 - 96 = 6).

Max trade % drawdown - The largest peak to valley percentage decline experienced in any single trade .

The above example expressed in percentage terms.

Max system drawdown - The largest peak to valley decline experienced in portfolio equity

This statisic is bit different, in the sense instead of taking a single ticker, it takes the equity as the ticker and provides the drawdown static.

Max system % drawdown - The largest peak to valley percentage decline experienced in portfolio equity.

Same as above expressed in percentage terms.

System Drawdown is useful to know since sometimes systems go into multuiple losses and the equity line plunges dramatically. For example, a system equity after reaching say 10,000 starts having loss trades (consecutively) and reaches 8000, this change would be shown here.

The system drawdown is a important factor to consider when deciding the capital requirement for the system, since if you are using leverage, a large system drawdown can ensure that you no longer can trade with the amount you have in hand and hence all permutations and calculations can go awry.

Benchmarking and Leverage - Prashanth K

A friend of mine was recently suggesting to a group, where I was part of, that the easiest way to pick stocks was to get MetaStock, run its indicators and select a list of stocks based on discretion.

I generally argue against such BS, but did not have the mood to do so and let it pass. But the information he sought to convey is that all you needed is MS and a data provider and well, you could well be on the way to riches and glory.

Just yesterday I was discussing with a friend (who is also a client of mine) about the importance of 'Benchmarking'. I believe if a system is not able to generate at the minimum twice the return of a said benchmark (after deducting all expenses incurred in regard to trading that system), it may be wiser to be a Buy and Hold (or shall I say Buy and Hope) investor - since one can spend the same time doing some other profitable work.

Generally when computing returns, people forget the risk they take (leverage) and instead calculate directly the net profit or loss. A friend of mine recently showed me a list showing the returns generated by him for his clients. He has over nine months averaged around 5 percent per month, net of brokerage and taxes.

While on the face of it, it's a commendable performance, what one misses is the fact that he uses nearly 5 times leverage to achieve such returns. The question that one should then ask is whether the risk is worth the reward. Five times the capital is no small leverage and one bad move can wipe out returns generated over months together. Hence once should carefully evaluate risk rewards before entering any kind of trade.

The author is a member of Bangalore Stock Exchange.