Is my methodology sound?

[jacknapier]

What I do is that I printed out 100s of charts and then backtest my strategy by simply placing a blank piece of paper over the chart and sliding the blank paper to the right revealing the chart as it moves. I then place the trades when I feel it has met my rules. Doing this my strategy is very profitable.

The problem is that it's a short strategy. And anyone experienced shorting knows that you can't just short any stock, your broker has to have that stock in inventory and inevitably whatever stock I want to play is never in inventory.

Anyway, I'm just curious, I want to turn this strategy upside down and try my method going long. But I just can't help but think that maybe it's not a good way to go about backtesting and that maybe I should invest money and time (both valuable commodities) into another software program? Does anyone have a better way?

 

[Giovan]

Jack get a programmer to code your systems. This should ease in backtesting/ However if you factor in the psych influence, you may realistically only gain 50% of the backtest result

 

[Purple Brain]

What I do is that I printed out 100s of charts and then backtest my strategy by simply placing a blank piece of paper over the chart and sliding the blank paper to the right revealing the chart as it moves. I then place the trades when I feel it has met my rules. Doing this my strategy is very profitable.

The problem is that it's a short strategy. And anyone experienced shorting knows that you can't just short any stock, your broker has to have that stock in inventory and inevitably whatever stock I want to play is never in inventory.

Anyway, I'm just curious, I want to turn this strategy upside down and try my method going long. But I just can't help but think that maybe it's not a good way to go about backtesting and that maybe I should invest money and time (both valuable commodities) into another software program? Does anyone have a better way?

If the process for backtesting the short side of your methodology worked well then I would say use precisely the same process for testing the long side. Nice to see someone using old tech rather than relying on computing power - has a solid work ethic feel to it.

Can I suggest you also start forward testing your methods on a demo account? As Giovan alludes to above, when you factor in the emotional aspects of trading - even demo - plus the additional workload of having to make a decision and do all your stop and position size calculations in real-time rather than having the comfort of making a pot of tea before you slide the piece of paper the next bar to to the right - your efficiency in converting back tested results into real ones will likely be less than 100%.

Bear in mind I am a fledgling trader myself, but what I lack in experience I have empathy in spades.

 

[NVP]

most Backtesting results display amnesia when forward testing.........

 

[malaguti]

What I do is that I printed out 100s of charts and then backtest my strategy by simply placing a blank piece of paper over the chart and sliding the blank paper to the right revealing the chart as it moves. I then place the trades when I feel it has met my rules. Doing this my strategy is very profitable.

The problem is that it's a short strategy. And anyone experienced shorting knows that you can't just short any stock, your broker has to have that stock in inventory and inevitably whatever stock I want to play is never in inventory.

Anyway, I'm just curious, I want to turn this strategy upside down and try my method going long. But I just can't help but think that maybe it's not a good way to go about backtesting and that maybe I should invest money and time (both valuable commodities) into another software program? Does anyone have a better way?

Hi Jack, Purple's advice is spot on.
There is just something I've picked up which could be right or wrong. when I read what you are saying, it seems to me that you want to use the SAME strategy in order to go long. Not necessarily the method of going through bar by bar.
If I'm right, then I would say its not a good idea to just use your strategy to go long. Often the characteristics of a market in freefall is often quite different to markets rising. They fall faster than they rise for one and also peaks are early and troughs are late is another. So my advice is use the same method of backtesting as you are..i wouldnt invest in another piece of software, i see its just wasted money and then when you think you have that long strategy (which will likely be different) thats where Purple has good advice in forward testing it on a demo account

 

[jacknapier]

Often the characteristics of a market in freefall is often quite different to markets rising. They fall faster than they rise for one and also peaks are early and troughs are late is another

Yea, I originally had in mind to measure buyer exhaustion. Which doesn't really translate well into selling exhaustion. When I have time tomorrow I'll rev up my printer and head to the local coffee shop and park for a few hours.

Thanks for all the advice here.

 

[jacknapier]

most Backtesting results display amnesia when forward testing.........

That it does. I had a strategy "tested" during the first half of the year, it was 100% accurate....until I tried forward testing it...

 

[new_trader]

That it does. I had a strategy "tested" during the first half of the year, it was 100% accurate....until I tried forward testing it...

There is a big difference between watching a bar on a chart as it is forming and seeing it after it has formed. The main flaw with back testing the way you have is that there is no way of knowing which came first, the high or the low, or how many times it touched either of them unless you have the time and sales data to accompany the chart.

 

[Hate2Lose]

What I do is that I printed out 100s of charts and then backtest my strategy by simply placing a blank piece of paper over the chart and sliding the blank paper to the right revealing the chart as it moves. I then place the trades when I feel it has met my rules. Doing this my strategy is very profitable.

The problem is that it's a short strategy. And anyone experienced shorting knows that you can't just short any stock, your broker has to have that stock in inventory and inevitably whatever stock I want to play is never in inventory.

Anyway, I'm just curious, I want to turn this strategy upside down and try my method going long. But I just can't help but think that maybe it's not a good way to go about backtesting and that maybe I should invest money and time (both valuable commodities) into another software program? Does anyone have a better way?

Trading requires work. There is no quick way. I spend all day scrolling through charts looking for entry points and making notes of the future. Your back-testing is interesting and by rights your methodology should work in reverse. If you keep calling right short, you should be ok the other way round. There is a trading adage that if you are continually unsuccessful trying to buy in at bottom of a falling market, to turn the chart upside down and carry on as you were! As you point out buying is easier because you don’t have the issue of borrowing stock. If your strategy worked for shorting do the same exercise for another 100 stocks (different to the first 100 – no cheating!). Time consuming but at least free (bar the paper and ink). Good luck, Let us know how you get on.