“A trade is
nothing more and nothing less than a datum point in a series of data points
subject to probability theory” – Peter Brandt
Conducting post trade analysis is imperative for
understanding the actual trading results of a system. Using a simple quadrant
can help categorize trades in order to glean information on yourself as a
trader which will help eliminate flaws. Every trade can be categorized in one
of four quadrants as follows. First,
“good trade, good result” implies that a trade was taken based on a pre-defined
trading signal in your business plan and resulted in a profit. These trades will create the right side of
the return distribution. Second, “good
trade, bad result” occurs when a trader takes a pre-determined trading signal
but is stopped out at a loss. For
particular strategies, like Trend Following, these trades will make up the bulk
of a return distribution but the losses (even in aggregate) are minimal when
compared to the total value of the gains. When using a Trend Following system,
traders let their profits run until presented with a reason to sell and cut
their losses quickly*.
The last two trade types occur when a trader deviates from
his or her plan. For a trader with a
defined strategy a “bad trade, good result” can occur when an impulsive trade
(i.e. a trade outside of the trading plan) is taken based on gut feel or some
other input. In most cases, the act of
“averaging down” by subjectively rationalizing the price’s action after the
trade is made and waiting for the position to turn a profit (or get back to
break-even) is a prime example of trades that fall into this category. Finally, a “bad trade, bad result” is simply
a trade that did not adhere to one’s trading plan and generated a loss.
For discretionary traders this quadrant below
will help identify weaknesses in your process. An assessment of trading results should be performed routinely. Tallying up trades into their respective quadrant will help you identify how well you are sticking to your trading plan and where improvement can be made on particular set ups. Moreover, identifying markets or trading environments that temp you to take "bad trades" and eliminating the urge going forward should improve trading P&L. Clearly, avoiding taking bad
trades is paramount to managing a robust trading system.
Good Trade
|
Bad Trade
|
|
Good Result
|
Trade that fits your system and generates a
profit
|
Trade that should not be in system but luckily creates a profit
|
Bad Result
|
Trade that conforms to your system but
generates a small loss, usually at the initial stop
|
Trade that is not allowed per system rules and produces a loss
|
In essence, traders should focus on process versus outcome.
You as a trader need to design, build, back-test, and execute on a pre-defined
trading plan. There should be no variation between the testing process and the
actual trading methods. Otherwise, you can expect to have random results
deviating from past expectations. Additionally, traders should have an intimate
understanding for the winning percentage of a system and how it will perform
during different market environments. This will aid in the ability to continue
to follow-through on the system’s rules in real-time trading. For example, a
trend trader would expect to pull less money, or even lose a little bit (i.e.
stop loss amounts), from a market with false breakouts. Conversely, when a
trend does emerge these traders need to continue to take system signals and be
ready to leverage into a positions once it begins to move in their favor. In
this case, whether a trend emerges and generates a profit or if a trade reverses
and the position is stopped out both should be considered good trades since
both adhere to the system parameters.
Peter Brandt adds, “There is a difference between a “good”
trade and a profitable trade. As a trader, your goal should be “good” trades.
Then let the profits land where they may over a multi-trade series.” Peter’s
words echo the perspective of focusing on process and not outcome. In doing so,
over a full market cycle your trading results should come to closely resemble
the back-tested data series and provide a better measuring stick for analyzing
your process.
*Trend Following by its nature captures outlier price moves
and the use of stop losses will build asymmetric return distributions with
right tails.
As always, please feel free to contact me with any
questions.
JD
Comments
Post a Comment