An Overview of
Backtesting and Paper Trading
As the sole owner and operator of a trading business, you play the
role of everyone from CFO to data entry clerk. You are also the
QA department. And like any company, if you want to be assured of
the quality of your product, you need to test it.
Unless you are detail oriented, you may not particularly enjoy the
role of tester. But testing is a crucial task to
perform if you want to give yourself – and make no mistake, you
are your own most
important client – a successful swing trading system.
Nevertheless, testing is
long and tedious process. It will try your patience. You will be
in the midst of testing, see positive results, and think: 'If only
I were trading for real, I'd be making money right now. I should
end these tests and get on with it already!' To be sure, the siren
call of live trading is strongest right after big profits are made
This is a temptation to resist at all costs. Do not cut your
testing short. There is simply no other activity that will
identify problems with your trading system as effectively as
running backtests and making paper trades.
During testing, for instance, you may find you are taking on too
much risk and need to adjust your trading plan to reflect the
harsh realities of the market. Or you may discover that your
intended basket of resources do not work well together and is in
need of compatible alternatives.
The component that most directly benefits from testing, however,
is the trading strategy. Often times testing reveals how a
strategy requires much more than a slight tweak to make it work.
Other times even drastic overhauls will not save it. At which
point you will have to abandon it outright and start again with a
clean slate. This is no doubt discouraging. But take comfort in
the fact that you may have just saved yourself from financial
There are qualitative benefits to testing as well. These are just
as important as those which quantitatively refine the profit
potential of your trading system.
for instance, is a great way to learn the ins, outs, and duration
of trading. Think of it as a free education where you
simultaneously gain valuable hands-on experience.
At a deeper level, testing poses a basic question: Do you have
what it takes to be a swing trader? If you are interested in
finding out sooner rather than later, make sure your testing
procedures mirror the way you will actually conduct your trades.
Then if you find you do not care much for trading after all, this
life lesson comes virtually free of charge.
But if you can discover the joys of swing trading even in the
testing phase, it will be much easier to keep your emotions in
check and resist the lure of live trading before you are ready.
After having spent so much time developing your trading plan,
trading strategy, and selecting your trading resources, you owe it
to yourself to thoroughly test how these components work together.
Doing so greatly increases the chances that your swing trading
system will be profitable once you expose it to the markets.
Backtesting is time consuming, especially for the discretionary
trader. Many backtesting sites have sprung up in recent years to
address this problem. Unfortunately, most of them only allow the
tester to work with a select set of testing parameters. While some
permit customization of their set, others do not. But even when
customization is possible, it is unlikely any one site will have
everything you need. Worse still, there are things in the trader's
universe that are not quantifiable at all. Take trend
, for example. Although much can be done to standardize
the way they are drawn, they nevertheless remain more of an art
than an exact science. So even though you may have programmed very
little wriggle room into your trading strategy, automated
backtesting sites may not be much use to you. This is doubly true
if your trading strategy is sufficiently complex so as to defy
What you should not
is construct a swing trading system around what these sites are
capable of testing. Why put yourself at the mercy of site
programmers when you do not have to? This defeats the purpose of
doing it yourself and generally dampens the entrepreneurial
There is a better way. Consider making backtesting an offshoot of
paper trading. There are many ways this can be done
But do not skip backtesting altogether. Why? Because you want some
idea how your trading system will behave in a variety of market
conditions, and not just in the one that currently exists today.
At minimum, you want to see if your system works in both bull and
bear markets. But what about in volatile and non-trending markets?
Or during seasonal and cyclical market patterns? Select some
well-defined periods of the past, run your tests, and find out.
Paper trading is live trading minus the actual risk. No real money
is at stake. You execute trades using a fake brokerage account.
Your wins and losses are
recorded 'on paper'
But does this mean you should go hog wild and trade whenever the
mood strikes you? Absolutely not. You must trade exactly
as you intend to trade when you go live with a real
account. Or as close to it as possible. Otherwise, you will not be
able to tell if your trading system works.
This means paper trade in real-time. You might be tempted to wait
until the end of the day or week, and then test retroactively. But
it is important to get a feel for the rhythm of trading. Paper
trading not only tests and refines the material components of your
trading system. It does the same for your skills as a swing
One reason why paper trading might be preferred to (but not a
substitute for) backtesting is because of its focus on the
present. Testing in today's market should have greater predictive
value for your trading system tomorrow than if all testing weight
was given over to the distant past. This is important because you
will soon be live trading. The rationale here is similar to what
drives some traders to prefer weighted and exponential moving
averages over the simpler, arithmetic version.
Just be sure to note how many trades you are making. On the one
hand, it is a problem if your system generates too few trades. You
do not want the funds in your account to idle in cash for long,
but to constantly turn over from one trade to the next. On the
other hand, if your system generates more trades than your real
account can handle, trade them all anyway on paper. For this is a
great opportunity. Not only will you quickly reach the magic
number needed to make your testing statistically significant (30+
trades minimum, 100+ better), you now have the luxury of being
choosy. Identify which trades are the best and what makes them the
best. Include those characteristics as part of your trading
strategy's selection criteria for future trades. Or else have them
be criteria that 'kick in' whenever the number of potential trades
exceeds the limits of your trading account.
Trading Resources Guide
Trading Documentation ⇨