This Is Like Having a Crystal Ball

Stock market, back testing

You’ve gotta love this time of year…

Summer’s hot humid days are a fading memory. In their place are crisp mornings, with lots of warm sunny days.

The climate isn’t the only thing I’m enjoying now. It’s also time for the Autumn Racing Carnival — an exciting mix of friends, fashion, and fun.

Horse racing has been in my blood for years. I wouldn’t be far off by saying that I’ve seen more derbies, cups, and slippers than some jockeys.

My fondest memory is of the mighty Kingston Town. I still remember cheering him down the Randwick straight as a youngster.

Last Saturday it was all on once again. And as usual, I was trackside to take in the action.

A highlight of carnival day is the feature race. I always watch it from a seat in the century old grandstand. It makes me feel like I’m a part of history every year.

Another aspect I enjoy is seeing how people bet. Some punters analyse the form, others put their faith in a ‘hot tip’ from a mate, while many wager on colours or a name.

And these strategies can get results…

Champagne corks were popping after each race. There was lots of backslapping and smiles. Winning tickets were coming out of wallets and purses around the course.

But here’s the thing…

Rarely did I see the same people celebrating.

Sure, there were lots of winners. But it was usually different people after each race.

I doubt many of Saturday’s betting strategies would stack up week after week.

The awakening

The whole betting culture got me thinking about trading.

You see, success has to be repeatable on a regular basis. Anything less is just a game of chance.

I gave up betting on the horses years ago. It turns out that I didn’t have an edge. And without an edge, you’re just giving money to those who do.

But it took me a long time to figure this out…

A few big wins would make me believe I knew what I was doing. I’d then lose money for weeks on end. I’d then wonder if my wins were just due to luck.

My awakening came during my university years. Frustrated by the swings in my TAB account, I decided to test my skills. I did this by keeping a record of every bet’s outcome for a year.

The results were revealing…

After countless hours studying the form, all I had was a small loss.

The jury was in.

I wasn’t going to be a professional punter. Race day has been a social event ever since.

So what does a day at the track have to do with trading?

Well, it made me reflect on my experiences.

It took me a year to confirm I wasn’t a skilful punter. That’s the nature of trial and error. You eventually get an answer, but you waste a lot of time in the process.

My experience as a trader is entirely different.

Gone are the days of a time hungry betting ledger. I now have a more efficient way to decide if a strategy has merit. It’s a process I believe every trader should understand.

Test before you trade

One of the best things about system trading is back-testing. This is where you use historical data to ‘road test’ a strategy. You can learn more in one afternoon than in a year of live trading.

But it must be done properly…

I often hear people talk about back-testing their trading ideas. While they may have the right intention, their methods for testing a strategy mostly fall short.

Back-testing requires lots of data. For instance, if you record a stock’s daily open, high, low, close and volume, that’s around 1,300 pieces of data per year. And that’s just for one stock!

Many people try to get around this by shortening their testing period. Rather than analysing a strategy over many years, they’ll analyse results over a few months.

But this is a mistake…

You see, the shorter the testing period, the greater the influence of chance. This could see you commit to a weak strategy after a lucky streak, or discard a good system due to some losses.

I avoid this pitfall by back-testing over not just years, but decades. This is how I gain confidence in a system’s ability to make money in live trading.

Take Quant Trader for instance…

You’ll probably be familiar with the system’s back-testing. From 1993–2014, the average winning trade was 37.9% and the average loss was 14.9%.

Now, these are good numbers in my books.

But I take back-testing further still…

I want to see the ‘average’ result occurring over and over again. This minimises the risk that one or two strong periods have had an undue influence.

Check this out:

Win rate Average profit Average loss Dollars won to dollars lost ratio
1993 to 1997 58.6% 32.7% 13.5% $3.42 to 1
1998 to 2002 49.4% 38.1% 15.4% $2.42 to 1
2003 to 2007 63.6% 35.2% 11.7% $5.25 to 1
2008 to 2012 51.4% 37.7% 17.7% $2.25 to 1
2012 to 2017 56.2% 32.2% 13.3% $3.11 to 1
Average 55.8% 35.2% 14.3% $3.29 to 1

 

The table shows Quant Trader’s hypothetical performance in five-year intervals. It invests $1,000 in every long trade. And as always, there’s no allowance for costs or dividends.

Now, what makes this so interesting is that each testing period is independent. A strong bullish phase in one five-year interval has no bearing on the next set of results.

Compare the results for yourself…

You’ll see the best interval was 2003–07, and the worst was 2008–12. The big respective events during these periods were the mining boom and the Global Financial Crisis.

Despite vastly different conditions, there’s still a similarity between the best and worst results. The same can be said for each of the other five-year periods.

I believe interval testing is an excellent guide to a strategy’s sustainability. It gives me confidence that the overall results are not due to a couple of super stocks or a lucky start date.

There’s one more thing I want to show you. I’m going to plot the performance of each five-year period on a graph.

Have a look at this:

Performance Chart

[Click to open new window]

This is what I call repeatable success.

The system has been profitable across all five-year intervals — despite their unique nature.

Back-testing is one of the most powerful trading tools I know. It can put a strategy through the wringer without placing a single trade. This could save you years of trial and error.

Knowing the history of your strategy also builds confidence. This is most important during the downturns. All too often, fear gets the better of people and they quit trading.

A bet on the horses can be a bit of fun. You may even be lucky and back a winner or two.

But if you’re serious (about trading or horses), then test your strategy first.

The key is being able to repeat success. This separates the pros from the punters.

Until next week,

Jason McIntosh,
Editor, Quant Trader

Editor’s note: Knowing a trading strategy’s past performance is vital. Not only can this save a lot of trial and error, it can also give you the confidence to trade it effectively.

Back-testing is a key part of the Quant Trader process. The aim of this rigourous process is to give you a statistical edge over the punters. Learn how to use it to your advantage here.

Seriously, this could change the way you trade forever. Check it out today.

There are countless examples of up-and-coming stocks beating the majors. But sadly, many traders only hear about them after the battle is won.

But that doesn’t have to be you…

Jason McIntosh’s Quant Trader is a fully algorithmic trading system. And it has a proven ability to detect big trends in ASX stocks early.

Check it out today. It could change the way you trade forever.

All graphics produced by Quant Trader unless otherwise noted.

Leave a Reply

Be the First to Comment!

  Subscribe  
Notify of