After some back testing, I noticed that the performance of most the pre-programmed strategies lag behind the benchmark "Buy and hold strategy". In addition to the effort of finding a good strategy that could beat the benchmark, I am curious how the buy and hold strategy works ? Especially about its position sizing management? For example, if I use portfolio simulation mode $100,000, with % of equity 7%, which means it will use 7% of portfolio to buy a stock when the entry signal appears. Let's say my data set has 100 stocks. But how does the "Buy and Hold" work? Does it use 7% or 1% of the portfolio to evenly buy all the stocks from the beginning, and last till the end ? I couldn't find any detailed information on all the documents or internet.

If the buy and hold strategy can beat 95% of other more complicated strategies, then why do people make more troubles rather than simply adopt buy and hold?

Could you help clarify ? Thanks.

QUOTE:
But how does the "Buy and Hold" work?

The Buy&Hold behavior is documented in detail in the User Guide at Preferences > Performance Visualizers > Performance. Please scroll down to the bottom.

QUOTE:
If the buy and hold strategy can beat 95% of other more complicated strategies, then why do people make more troubles rather than simply adopt buy and hold?

There are several reasons. Some can't sit tight and need action - not everyone's comfortable with passive investing. Many can't stomach the drawdown associated with B&H - be it the depth, duration or both. So the "trouble" you're referring to is an attempt to achieve a comparable reward for a much lower risk. And the more time exposure you have the greater the risk of a black swan event - like when you buy or short on 9/10/2001 and become allowed to exit only on 9/17/2011.

I found the B&H position sizing info. Thanks.

And the documentation says, if a trading system can achieve higher profit per bar (more efficient) with less exposure and less risk than B&H, we can increase the position sizing for that strategy to get higher return, with less risk, which makes sense.

Thank you.