Drawdowns

[howardbandy]

Greetings --

AFL code:

// techtrader v2 amibroker version
// here we define buy conditions and name each one as a variable
PositionSize = -10; // always invest only 10% of the current Equity
cond1=Cross(H,Ref(HHV(H,10),-1)); // when todays high crosses last highest high over the last 10 periods
cond2=H > EMA(C,40); // todays high is greater than the 40 day Exp MA of closes
cond3=HHVBars(H,70) == 0; // todays high is the highest for 70 periods
cond4=EMA(V*C,21) > 500000; // ensure at least $500k of money flow
cond5=C < 10.00; // only trading in stocks less than $10
cond6=C > O; // todays close higher than open

// the following line is the trigger if all conditions satisfied
Buy=cond1 AND cond2 AND cond3 AND cond4 AND cond5 AND cond6;

// here we define variables used once in the trade
ApplyStop( stopTypeLoss, stopModePercent, amount=10 );
Sell= Cross(Ref(EMA(L,180),-1),C); // close crosses below yesterdays average of the low
// here we define what gets displayed on the chart
shape = Buy * shapeUpArrow + Sell * shapeDownArrow;
PlotShapes( shape, IIf( Buy , colorYellow, colorRed ), 0, IIf( Buy , Low, High));

Filter = Buy; // lists exploration results conforming to our buy criteria
AddColumn(Buy, "buy", 1.0); //

Why does it work??

The simple reason is that the broad market, and most of the stocks in it, have been rising strongly for the past 25 years.

This system restricts purchases to stocks with prices below $10. I made a posting to one of the threads here at ASF about historical prices, and how splits, distributions, and dividends distort historical prices. The point is, you cannot trust that the price in the historical data base was the price the stock actually traded at.

I ran the system on the S&P 500 stocks beginning 1/1/1995.

The system bought ebay on 11/23/98 at $8.04.

ebay did not trade at $8.04 in 1998. There have been four splits, totaling 24:1.
http://investor.ebay.com/faq.cfm
ebay was about $192.96 on 11/23/1998.

It would be interesting to test over a period of market decline, but that would require either removing the $10 limit (and changing the system), or having unadjusted historical data.

When I changed the test dates to begin 1/1/2005, using the 500 stocks in the S&P 500, only 15 stocks had trades, 6 were winners, 9 were losers. (I did not check to see if all of the trades could have taken place at less than $10.) The total exposure of the resulting trades (all 500 stocks for all 2.7 years) is 0.00006 -- 0.006% of the time. That is not enough data to allow meaningful statistical tests. And it is not enough data to give me a subjective feeling of comfort.

So --- why does it work? At the risk of getting some flack back, the cheeky answer is -- the system is confusing brains with a bull market.

Thanks,
Howard

 

[theasxgorilla]

It would be interesting to test over a period of market decline, but that would require either removing the $10 limit (and changing the system), or having unadjusted historical data.

So the sub-$10 filter is a quick and easy filter to try and limit the universe to stocks that have the greatest growth potential. I think it's too low for the US market. Does such a psychological price level exist on US stocks?

Also, due to the relative size differences between the 2 markets if I were to apply TT on the US I reckon you'd want to include some of the Russell 2000 stocks. On the ASX companies from 100th to 300th by market cap are categorically small caps, but you can trade them on margin. Does such a group exist in the US?

I agree 100% with what you describe re: unadjusted historical data. I can't see a way to solve the dividend problem, but I think we're in a somewhat fortunate position here on the ASX as stock splits occur seldom, compared with the US.

Take QBE for example...back in 1985 it was trading at just over 50c. Now its up over $34. From what I can tell it's never split. In the US you might say don't buy stocks under $10 or $15 USD because they're there for a reason ie. That level in Aust (as per an estimate by Nick) could be as low as 20c AUD. Even if it's 50c, that means in US market terms QBE is currently trading at $1,170...if 20c is that level then QBE is trading at just under $3,000 a share!

 

[tech/a]

So --- why does it work? At the risk of getting some flack back, the cheeky answer is -- the system is confusing brains with a bull market.

You know I'm happy to wear that.
I managed to trade the bull market and do OK with no brains.
A few others with no brains either also did well.Stevo's method would no doubt come under a similar umbrella.

On the other side I know quite a few with plenty of "Brains" who failed dismally.

It would be interesting to test over a period of market decline,

Howard why is it that youd like to test a method designed for trading long in bullish conditions in a prolonged bear market? Its not designed for that?

Howard with an average holding period of over 300 days No doubt your sample period of 2005 till now showed a poor result. Id be interested in the open profit in those trades still open at the end of the test period.

I also believe that any system that performed well in a bull or bear market could be easily explained away as succeeding ONLY due to market conditions.
It appears that the trick seem to be in systems design to design a method which works in conditions in which its not designed.

At the risk of again being labelled Egotistical again.
Market rises 100% in an un precidented Bull run.
Dumb system out performs that 10 fold.
I can live with that and the label.

I can actually afford to shout ASX a beer!

This is WHY it works---SIMPLE!

(1) It holds its winning trades Far far longer than its losing trades.
(2) It makes much much more from its winning trades than it does from its losing trades.

Whos the smarter.
He who has a degree or he who employes those with degrees?

 

[tech/a]

Perhaps I could be so "Cheeky" as to suggest that perhaps brains are being confused with potential profits.

I am really suprised that Both Howard and ASX have missed the ONLY reason this systems makes above average $$s.

Just wanted to highlight and seperate so that the poor unfortunates who dont have the acumen of some here can see that you really dont need it---well not all of it.

This is WHY it works---SIMPLE!

(1) It holds its winning trades Far far longer than its losing trades.
(2) It makes much much more from its winning trades than it does from its losing trades.

However for another couple of cartons each (Perhaps Bud would be more to Howards liking).I'm happy to be proven incorrect by our learned gentlemen.

 

[Nick Radge]

Howard,
On the contrary, the adjusted chart of our market since 1952 is nothing short of one way traffic. I do have the data back to 1875 but it just offers more of the same. There have been 3 generational bull markets since 1952:

1950 - 1969
1974 - 1987
1991 - ?

The other "common sense" assumption which may be out of the realms of academia is that companies naturally grow and with that growth comes share price increase. We're not talking commodity prices.

Perhaps the US markets are "mature" and have stopped this process and that Aust is still a "developing" market. However, as I stated in another post, I'd gladly take a random set of stocks from the US market and run my system over them.

 

[stevo]

The simple reason is that the broad market, and most of the stocks in it, have been rising strongly for the past 25 years.

There have been pockets of decline that are probably worthy of focus. 2002 was not the best year and there was a very rapid spike down in 2001. 2000 was also very volatile.

I started system testing at the end of 2001 and didn't really have a system that I was comfortable with until the beginning of 2003 - right at the start of an incredible bull market! So I was happy to take what the market gave.

If I made the system longer term and used monthly charts I got great returns but was uneasy because the number of trades taken dropped in the system test and the sample size reduced. Data quality is also an issue - data gremlins are nearly impossible to avoid.

I am more comfortable testing over a large number of stocks 500 to 1000 ideally over a 10 year period. Testing just 20 or 30 stocks, or even 100 makes me quite uncomfortable. I like to think that I am not testing on 500 stocks rather just one stock (500 stocks joined end to end) over a very long time.

System testing will only give us a rough approximation of a possible future. If we are in a bull market then why not take advantage of it, fully aware that market conditions change. It is possible to go to cash if the markets change - something that has happened to me a couple of times over the last 5 years.

Trying to buck the long term trend by shorting stocks is like trying to swim against the rip at North Avalon beach - I stop swimming and wait.

stevo

 

[theasxgorilla]

This is WHY it works---SIMPLE!

(1) It holds its winning trades Far far longer than its losing trades.
(2) It makes much much more from its winning trades than it does from its losing trades.

I think you should introduce Howard to Coopers, it's one of the better Aust beers IMO

Tech/a, you're right. I still feel there is one final ingredient...luck, or intuition. Far be it for anyone of us to kick a gift horse in the mouth...but by at least acknowledging fortunate timing you won't mislead anyone into thinking you hold the master key to riches and copies are available in TradeStation, Metastock and Amibroker format.

I wonder if we are now at a point where we seperate out the semantics of what we're discussing, as being a budding system designer myself I'm more interested in the HOWs.

For those that are willing to take advantage of the WHYs without the inclination or wherewithal to investigate the HOW, then I agree with Nick Radge that the best thing you can do is find somone who runs a sophisticated system, invest money with them and stay the course.

I'm personally wary of turning on a simple system like TT at this potentially late stage of a bull market...not without some 'insurances' coded in at least.

 

[stevo]

I'm personally wary of turning on a simple system like TT at this potentially late stage of a bull market...not without some 'insurances' coded in at least.

Startup can be quite tough for a long term system since the system hasn't got any winning trades running and can take a while to latch on to some.

Have a look at the first 12 months of the closed trade equity curve - and then walk it forward 6 months at time from 2000 onwards to see how it performs on startup.

stevo

 

[tech/a]

I think you should introduce Howard to Coopers, it's one of the better Aust beers IMO

Tech/a, you're right. I still feel there is one final ingredient...luck, or intuition. Far be it for anyone of us to kick a gift horse in the mouth...but by at least acknowledging fortunate timing you won't mislead anyone into thinking you hold the master key to riches and copies are available in TradeStation, Metastock and Amibroker format.

You could call any longterm investment in a business "Luck" when it finally hits.If you place your self in the position with the right tools its amazing how luck will find you.

*Trading a long system through not so favorable times to find yourself in a new bull market
*Buying houses and sitting on them for years while others tell you your an idiot,to then take part in a housing boom doubling and trebeling sale prices.
*Working within your business to gain maintain and expand your edge after years finding yourself at the very top and securing the larger majority of business within your field.

Luck--??

I wonder if we are now at a point where we seperate out the semantics of what we're discussing, as being a budding system designer myself I'm more interested in the HOWs.

For those that are willing to take advantage of the WHYs without the inclination or wherewithal to investigate the HOW, then I agree with Nick Radge that the best thing you can do is find somone who runs a sophisticated system, invest money with them and stay the course.

ASX your not alone this is where all start but never end their journey or more to the point get past.
Unless you know WHY and there are only 2 others I know of (WHY'S)---you'll be for ever twiddling and tweeking the HOW.

I'm personally wary of turning on a simple system like TT at this potentially late stage of a bull market...not without some 'insurances' coded in at least.

Again not alone.
Often looking at individual stock in a portfolio gives you the same feeling.How can this keep going,the temptation to sellout is huge.The chart below could be any stock and any index.The comments apply to any chart which looks similar.