All Systems Are Predictive

[DionysusToast]

Seems to me, we still have some believers in trading without having to have the ability to make a call on what will happen in the market you trade.

First of all, consider that the markets are like a bucket of water. You can put water into a bucket but it is impossible to take out more water than you put in. Similarly, the markets have a finite amount of money in them and only money that has gone in can come out.

This is relevant because it implies a lot about system degradation. There cannot be a system that will work for everyone. It is impossible. Let's say some super system came out that made everyone using it a winner. If all market participants eventually used it, then everyone would be able to take more money out of the markets than they put in. This is obviously impossible. The system would degrade because market conditions would adjust to it. There is no other alternative.

If market conditions MUST change to make all system fail with overuse, then all systems are by definition reliant on prevailing market conditions. It doesn't matter how you slice it, all systems that work do so because they predict that market conditions will be complimentary for the duration of the trade.

So - whilst it's lovely to hear about all sort of magical systems using special techniques, they are all fundamentally predictive.

Consider this next time someone tells you to toss a coin for entry.
- What market conditions does the overall system need to be profitable ? Any market conditions at all ?

Impossible.

 

[HowardCohodas]

Seems to me, we still have some believers in trading without having to have the ability to make a call on what will happen in the market you trade.

Count me in that group.

 

[Trader333]

There cannot be a system that will work for everyone

Well in my view it is possible but only theoretically and the stock market is largely built on that belief. However, it does require constant economic growth and an ever expanding population to sell stocks to that have gone up in value and then the next generation would sell on again etc.

Practically though I agree with you.


Paul

 

[DionysusToast]

Count me in that group.

OK then, fair comment.

Perhaps you can give us an example of a type of trade that once placed, will make money regardless of what happens in the market you are trading. I will then explain to you the conditions in which it will fail and the conditions in which it wont.

Please - don't resort to "that'd reveal my secret edge" - it's too early in the conversation for silliness.

 

[Splitlink]

There's something here that I don't understand, DT, and that is your argument that markets are finite.

Well, they must be at some point, but I don't see how that has any bearing on the likes of us.
You could say that about oxygen, but I don't regard that as an immediate problem.

What I see about systems is that, separating the scams, no matter how successful you, as an inventor of a system can be, you will not be successful in imparting your knowledge to the majority of students. Ihat is as impossible as it is to beat the market.

There might be a very few students, who are little more than clones, who may be able to copy your system. But anyone with any individuality in them, at all, will differ in some minute way from the system.

Unfortunately, they could well be the first to cry "SCAM" when, in fact, the fault lies with them.

 

[HowardCohodas]

OK then, fair comment.

Perhaps you can give us an example of a type of trade that once placed, will make money regardless of what happens in the market you are trading. I will then explain to you the conditions in which it will fail and the conditions in which it wont.

Please - don't resort to "that'd reveal my secret edge" - it's too early in the conversation for silliness.

As you wish.

Random Entry:
Random entry experiments and anecdotes of those who claim to practice random entry lead me to believe that predicting the direction of the trading instrument is unnecessary to being profitable as long as you practice effective money management.

This result depends on markets continuing to trend. Evidence is that they have and continue to trend around 20% of the time. I see no reason to believe that this will not continue.

Selling time
Here we could quibble semantics so I'll try to be careful.

I trade index options credit spreads and form an iron condor when market conditions permit. I do not make a prediction on the direction of the market when entering a trade. In fact, I do not even look at the chart of the underlying instrument. I do predict that time will pass.

Here comes the quibble part. My entry rules are based on probability of the underlying instrument touching the short option strike price. The second entry criteria is that I get sufficient credit for the risk level (represented by the probability of touching) I set.

I've afforded you three points where you could ague that I do, in fact, predict the market direction even if indirectly. The first is that the calculation of probability of touching, which is primarily derived from time and volatility, has a built in consideration of predicting direction. Then the argument comes down to "is the predictive component a primary, secondary or tertiary one."

The second point where you might argue that predicting direction is a component is my choice of "sufficient credit." The credit received in a credit spread does have market direction sentiment built in to the prices buyers are willing to pay.

And the third is really a combination of the first two. The strategy is most profitable when an iron condor can be formed. Since "market conditions" dictate the effectiveness of doing so, then one could argue I am making a prediction by choosing whether the conditions are right to form the iron condor.

Hidden reveal
Please point out where I have been less than forthright in answering all questions about my strategy or where I have resorted to the "secret edge" to hide anything. Please don't tell me that you are using the tentative title of my book as the basis for your remarks.

 

[HowardCohodas]

But anyone with any individuality in them, at all, will differ in some minute way from the system.

I resemble that remark. Anyone who became an engineer probably has tinkering in their DNA. I suspect that could also be said of a trader.

Part of good strategy design is an analysis of robustness. Robustness has many definitions, but for purposes here I'll use the idea of sensitivity analysis. How sensitive is the strategy outcome to market conditions? How sensitive is the strategy outcome to changing parameter settings? Is there a discontinuity built into the model where small changes in market conditions or parameter settings make small changes in outcomes but there is a hidden cliff where the next small increment of change results in total disaster.

Robustness testing of a strategy does not guarantee success, it is just one more layer of qualification that weeds out the weak strategies.

 

[Jaydee]

Slight tangent, but I hate that word 'predicting'.

For example, say I am delt a pocket pair of Aces in No Limit Texas Hold'em Poker. Against one other player with a random hand of two cards, before any other cards are seen, I am roughly an 83% favourite to win the hand. From this, if I get my money in the pot against the other guy, I am not predicting I am going to win - I'm just saying 83 times out of 100 I am going to win, over an infinite sample size of hands.

It's the same for trading is it not?

 

[Splitlink]

I resemble that remark. Anyone who became an engineer probably has tinkering in their DNA. I suspect that could also be said of a trader.

Part of good strategy design is an analysis of robustness. Robustness has many definitions, but for purposes here I'll use the idea of sensitivity analysis. How sensitive is the strategy outcome to market conditions? How sensitive is the strategy outcome to changing parameter settings? Is there a discontinuity built into the model where small changes in market conditions or parameter settings make small changes in outcomes but there is a hidden cliff where the next small increment of change results in total disaster.

Robustness testing of a strategy does not guarantee success, it is just one more layer of qualification that weeds out the weak strategies.

That was not the point that I was trying to make, Howard.

If you are the originator of a strategy, no matter how robust, or good, or how your DNA is made up, I maintain that you will not be able to impart your knowledge, exactly, to the majority of your students. In some, perhaps microscopic, way there will be a difference.

That difference may make it better or worse, but it will no longer be your system, but an adaptation.

 

[HowardCohodas]

That was not the point that I was trying to make, Howard.

If you are the originator of a strategy, no matter how robust, or good, or how your DNA is made up, I maintain that you will not be able to impart your knowledge, exactly, to the majority of your students. In some, perhaps microscopic, way there will be a difference.

That difference may make it better or worse, but it will no longer be your system, but an adaptation.

I got so carried away with my answer, I forgot to connect it to the question. My lame excuse is jet lag, even thought I'm only 3 time zones away from home.

The connection I was trying to make is that the likelihood of a student being successful with a strategy invented by others is enhanced if a sensitivity analysis shows the strategy is robust and outcomes do not vary much based on market changes and parameter tweaks.

Now rule/logic changes is a whole other matter.

 

[Splitlink]

I'm beginning to think that I may be getting jet lag, myself. I hope It's not contagious.

 

[Shakone]

Random Game A

Suppose we have a game where we toss a 'coin A', if we call it right, we get £10, if we're wrong we lose £10. Suppose also that the probability of heads is 0.6 (60%). Then we can have a strategy with positive expectation, by betting on heads. Even if we charge say £0.50 to play this game, then we can still have a positive edge. Nothing hard to understand there. What we required was that probabilities weren't 50-50 (sufficiently different from that in fact), that we also knew which side had the higher probability and that the win size was large enough in comparison to the cost to play/spread (0.50p).

Now suppose we could play this game twice in a row (stage I and stage II), but it is our choice whether to play stage II, with the condition we have to keep the same side (if we were long/heads in stage I, we have to be long/heads in stage II). Again we'd be able to have a positive edge by going for heads in stage I. Again nothing unexpected there I hope.

Random Game B

Finally, suppose we have another coin which is fair i.e. 50-50. And we toss this coin, and whatever comes up, that is what we have to choose for stage I in the previous 'random game A' with that 60-40 coin (still our choice to continue into stage II). Can we ever have positive expectancy from this game? The answer is yes, isn't it? What will it depend upon? The same things....that the probability wasn't 50-50, that the win size was large enough in comparison to spread, and that we knew which side was more likely to come up.

But this second game was random entry. So it should be clear now what market conditions you need to make money from it.

 

[meanreversion]

If market conditions MUST change to make all system fail with overuse, then all systems are by definition reliant on prevailing market conditions. It doesn't matter how you slice it, all systems that work do so because they predict that market conditions will be complimentary for the duration of the trade.

So - whilst it's lovely to hear about all sort of magical systems using special techniques, they are all fundamentally predictive.

Consider this next time someone tells you to toss a coin for entry.
- What market conditions does the overall system need to be profitable ? Any market conditions at all ?

Impossible.

Is the point here that a system needs the right conditions to generate profits at that time? Fully agree with that. A trend system needs trends, otherwise when the market is chopping around, the account equity will not be increasing, at that time.

However, what I think you're skipping here is this - we know markets are not trending most of the time, but how about if during the short time that they ARE trending, the profits accumulated more than compensate for the flat/small negative returns when the market is not trending?

As an aside, everyone is an agreement that there is no "perfect system", because if it existed, everyone would use it blah blah. But one man's perfect system is another man's poison. I came to the realisation recently that most beginners looking for a system to trade struggle to find one, simply because they have no basis for comparison - a metric to judge systems, if you will.

You can see it from some of the comments on this forum, remarks about making 50% a month and so on. If your basis for comparing systems is raw profit (as this sort of comment would imply), you'll come unstuck fairly quickly.

 

[DionysusToast]

Well in my view it is possible but only theoretically and the stock market is largely built on that belief. However, it does require constant economic growth and an ever expanding population to sell stocks to that have gone up in value and then the next generation would sell on again etc.
Practically though I agree with you.
Paul

Actually - it is only by misunderstanding the markets that you could say that this was theoretically possible. As was stated, the markets are like a bucket. Money goes in, more money cannot come out.

Consider a stock. Let’s say a stock has 30 million shares. It started at $1 and is now at $100. There is no way all buyers of this stock can make money. There are 2 reasons. First of all, if all holders sell, it is IMPOSSIBLE for them all to get the $100 unless there is a buyout. The number of shares * price = total value is a myth. If all people sell, then the majority of sales will be made on the way from $100 down to $0. The act of selling lowers price. This is as it should be. Also – the act of selling creates a buyer that also needs to profit for all participants to make money.

Bottom line is – all people who put money into the markets cannot make money. Only what has gone in can come out.

There's something here that I don't understand, DT, and that is your argument that markets are finite.

Well, they must be at some point, but I don't see how that has any bearing on the likes of us.
You could say that about oxygen, but I don't regard that as an immediate problem.

You are right – these concepts only come into bearing when people come onto forums and claim to have the trading equivalent of a perpetual motion engine.

As you wish.

Random Entry:
Random entry experiments and anecdotes of those who claim to practice random entry lead me to believe that predicting the direction of the trading instrument is unnecessary to being profitable as long as you practice effective money management.

Howard – you were asked to be specific. The random entry myth will be debunked here if you provide the details of that system. On other threads you claim to only believe proper scientific studies but in this one – anecdotal evidence “leads you to believe”…

As you wish.
Selling time
Here we could quibble semantics so I'll try to be careful.
I trade index options credit spreads and form an iron condor when market conditions permit. I do not make a prediction on the direction of the market when entering a trade. In fact, I do not even look at the chart of the underlying instrument. I do predict that time will pass.

You and I both know this is predictive. You are selling time BUT there are prices at which you will lose money. Your Iron Condor makes money if the price stays within a certain range. So – by placing such a trade, you are predicting that the price will stay within that range. This is a predictive strategy.

“Max Loss Occurs When Price of Underlying >= Strike Price of Long Call OR Price of Underlying <= Strike Price of Long Put”

This is making a prediction Howard. To claim otherwise would be ludicrous.

Whether you are making money from lack of volatility, lots of volatility, direction or whatever - you are still predicting that will happen.

Too many people believe they don't have to make a call...

 

[meanreversion]

Iron Condor man?

http://www.google.co.uk/imgres?imgu...:429,r:10,s:109&tx=166&ty=33&biw=1440&bih=839

 

[meanreversion]

Howard, if I were you, I wouldn't bother arguing with DT.

1. He knows a lot more than you
2. He's stubborn and won't let things rest

 

[meanreversion]

Too many people believe they don't have to make a call...

The more I think about it, the more I come to the conclusion that separating traders into "discretionary" and "mechanical" is bogus.

A discretionary trader will use rules to guide his trading - money management is key, after all.

A mechanical trader uses discretion to select his markets, the parameters, the bet size and so on. He will also retain discretion over adjusting the system on occasion.

You have to make a call, whether it's the markets you trade, the direction you think they might move, the relation between implied and actual volatility, etc.

 

[DionysusToast]

The more I think about it, the more I come to the conclusion that separating traders into "discretionary" and "mechanical" is bogus.

A discretionary trader will use rules to guide his trading - money management is key, after all.

A mechanical trader uses discretion to select his markets, the parameters, the bet size and so on. He will also retain discretion over adjusting the system on occasion.

You have to make a call, whether it's the markets you trade, the direction you think they might move, the relation between implied and actual volatility, etc.

I agree with you. There's a third group though - they are the seekers.

This thread is mostly for them.

If they can understand the bucket. They can understand that magic bullets only cause leaks.

I like the idea of people embracing the decision making process and having a feedback loop (i.e. journal of sorts) to make improvements in that decision making process. I just feel a little sad for the seekers of systems who spend their time seeking and bleeding money.

 

[oildaytrader]

The more I think about it, the more I come to the conclusion that separating traders into "discretionary" and "mechanical" is bogus.

A discretionary trader will use rules to guide his trading - money management is key, after all.

A mechanical trader uses discretion to select his markets, the parameters, the bet size and so on. He will also retain discretion over adjusting the system on occasion.

You have to make a call, whether it's the markets you trade, the direction you think they might move, the relation between implied and actual volatility, etc.

So you live in a world of make believe about system traders.Override systems because you believe in discretion to select markets, if selection is wrong then reselect,and if wrong again reselect etc etc etc etc etc etc

Don't know why this complete novice has no clue.

I never change automated systems ,nor do I interfere with them.They are good systems, therefore your systems must be real crap to need changing,tweaking and optimising for different conditions, and discretion needs to be applied regularly.

I don't know why people bother with forums novices.

 

[DionysusToast]

2. He's stubborn and won't let things rest

No I'm not....