all public
systems are run daily using a standard set of parameters - symbols,
start and end date, position sizing, etc.
to show
the effect of slippage and commission on performance, each system is run once for each of the following combination of slippage and commission
parameters:
- no commission or slippage (raw profit)
- commission set to 5 one way, no
slippage
- no commission, slippage set to 0.2%
- commission set to 5 one way, slippage set to 0.2%.
the initial
capital is set to 100,000 and entry size is set to 10,000 for all trades.
various statistics are shown, along
with the "Tradery score" which is calculated using the following
formula:
APG * ( 1 - signAPG * EXP/100) * ( 1 - signAPG * max( UI, 20 ) / 20 )
where
- APG - Annualized Percentage Gain
- EXP - Percentage exposure
- UI - Ulcer Index.
- signAPG - +1 if APG > 0 or -1 otherwise.
In determining this formula, the initial requirements were:
- the score should be + for a gain, - for a loss, and 0 for a system
that doesn't generate positions
- the values in the formula should give a global measure of the value
of the system including raw performance (APG), risk (EXP) and
tradability (UI).
- the higher the score the better the system
- the values used in the formula should be preferably uncorrelated
Users' comments are welcome and will be taken into account when fine
tuning this score calculation formula in future releases.
results of
the automated daily runs are
affected by the built in equity and position sizing logic, which are
required for a realistic simulation. For example, positions that would cause
the amount of available cash to become negative are not taken. This may even
lead to seemingly inconsistent results, for example a system may show a
better return with commission than without. This is simply because one or
more trades that were entered during "without commission" testing, could not
be entered any more in "with commission" mode because they would exceed the
amount of available cash, and were replaced with one or more other trades
that had a better gain, overall more than compensating for the loss due to
the commission fee.
