Market Wizards - Survivorship Bias ?

[DionysusToast]

Ok - let me throw out a 'Talib-esque' hypoethtical theory for those that don't know it.

Let's say you have a bunch of coin toss traders.

If you take 10,000 traders, at the end of the year, perhaps 5,000 will beaten the market just by luck
At the end of the 2nd year, 2,500 of the 5,000 will have got lucky again
At the end of the 3rd year, 1,250 will have beaten the market again
year 4 - 625
year 5 - 312
year 6 - 156
year 7 - 78
year 8 - 38
year 9 - we have enough people to write a Jack Schwager book.

Let's face it - no-one will be interested in the 9,981 people that didn't have the 9 year run of luck. Books about people that lost half their account in 6 months do not fill the shelves.

Let's also face facts that just because of the sheer number of people trading, their will inevitably be those that trade like Spanish but get away with it for years.

Now - if you pick those 19 people left, that just had a run of good luck, is there any reason to think that their luck will be any better than anyone else on the planet ?

I would say not.

Now - Jack has updated the original Wizard books & in the 2nd editions he goes back to those he interviewed and a good number of them started to lose money or barely break even. This is what you would expect if their previous run was more luck than judgement. Of course, they all have reasons for their runs of bad luck.

Would it be fair then to say that a number of the wizards in these books are not wizards after all - they are just a statistical anomaly ?

 

[dcraig1]

Survivorship Bias ? Possibly part of the story. I've read Market Wizards and New Market Wizards and I didn't find either particularly interesting.

Possibly the thing common to many of the interviews is that many claimed that they come to understand and control risk. If you can truly do that, your chance of being once of the survivors is greatly improved.

 

[Vaco]

you have to control the risk to your capital long enough for your good run to come along and make you some money. Thats why money management is so important, so that a run of losing trades does not knock you out.

All things are not equal in trading sooooo many variables between one trader and another but I still believe the key one as to how long they survive is control of risk.

 

[DionysusToast]

Maybe my point has been missed somewhat.

Does anyone think that a bunch of people in Schwagers books may have got there through sheer blind luck ?

 

[Vaco]

no, otherwise they would not be on here trying to replicate that success.

 

[A Dashing Blade]

No. Careful reading of the book shows that most of the interviewees share very similar trading characteristics and psychologies

 

[Martinghoul]

To me, it's a given that business literature, especially its better, more famous examples, is full of survivorship bias. So the phenomenon is not confined to trading literature per se.

 

[timsk]

Does anyone think that a bunch of people in Schwagers books may have got there through sheer blind luck ?

Hi DT,
It's possible I suppose, but not very probable. Your premise wouldn't really make sense when applied to a day trader doing five trades a day or more. (I forget whether or not any of the wizards are day traders?) The chances of them being profitable every week for ten years must make the odds of winning the National Lottery look positively attractive. So, I don't subscribe to the sheer blind luck theory in their case.

A more likely explanation, IMO, is that the ones who have fallen by the wayside and are no longer profitable - changed their trading style in an effort to adapt to changing market conditions. They got it right once, but can't get it right anymore. Add to that the fact that anyone who trades for ten years is a different trader in year ten to the one they were in year one. So, I doubt that any of them were successful purely through sheer blind luck alone and those that have managed to maintain their success have done so by managing to evolve and adapt their style successfully to an ever changing market.
Tim.

 

[donaldduke]

Does anyone think that a bunch of people in Schwagers books may have got there through sheer blind luck ?

If if it was talented coin syndrome, all, every single one, 20 years later would have failed by now.

Also, alot of them were floor traders and they going struggle in the electronic age.

Just like if you took most profitable electronic traders today and threw them back onto the floor most would probably fail, i know i would.

So i think more of issue of having outdated skills/not being able to adapt rather what they were doing previously was just luck.

 

[DionysusToast]

If if it was talented coin syndrome, all, every single one, 20 years later would have failed by now.

Well - that's my point.

In the 2nd (or 3rd - it's not with me here) edition of the first book, Schwager updated it with follow up interviews with the same traders a few years later.

In each case, the trader gives an excuse as to why their trading suddenly faltered. Not a single one of them considered that their luck may have simply ran out.

A number of them have indeed been failing since the original interview, which you would expect to be the case if lady luck had indeed played a part.

I find it interesting that some people can't digest the fact that sheer mass of people in the industry combined with the laws of probability means that traders like this will exist through luck alone - even for 20 years.

Considering how much some people value 'probability' when it comes to their own trading, it is most interesting how difficult it is for them to apply the simple concept when it comes to people they aspire to be like.

 

[donaldduke]

Well - that's my point.

In the 2nd (or 3rd - it's not with me here) edition of the first book, Schwager updated it with follow up interviews with the same traders a few years later.

There is no such follow up interview book that i know of. Do you have a link?

Not for the first two books anyway, there may have been one for his third book which i dont own and i think contained interviews with stock traders near the height of the internet boom.

 

[DionysusToast]

Hi DT,
It's possible I suppose, but not very probable. Your premise wouldn't really make sense when applied to a day trader doing five trades a day or more. (I forget whether or not any of the wizards are day traders?) The chances of them being profitable every week for ten years must make the odds of winning the National Lottery look positively attractive. So, I don't subscribe to the sheer blind luck theory in their case.

A more likely explanation, IMO, is that the ones who have fallen by the wayside and are no longer profitable - changed their trading style in an effort to adapt to changing market conditions. They got it right once, but can't get it right anymore. Add to that the fact that anyone who trades for ten years is a different trader in year ten to the one they were in year one. So, I doubt that any of them were successful purely through sheer blind luck alone and those that have managed to maintain their success have done so by managing to evolve and adapt their style successfully to an ever changing market.
Tim.

Tim

Your more likely explanation is not actually more likely at all. How do you quantify that your explanation is more likely than any other ?

I could certainly go to lengths to show the probability of traders like this evolving through sheer chance but Taleb has done that job already.

To some extent, I would imagine that most would-be traders would not want to entertain this possibility for a second.

Imagine if half of the people in Schwagers book weren't that good after all - just statistical anomalies. How would the aspiring trader be able to handle such information about their heroes and still battle on in their attempts to trade succesfully themselves ?

On a purely intellectual level, there is every chance that this is a possibility and someone with emotional detachment to the subject would be able to accept it as a possibility.

Of course, the fact that we have some are emotionally attached to this topic means it's very hard for people to accept the possibility.

 

[D70]

Hey DionysusToast,
You are pretty good at raising interesting threads! I take my hat off to you.
When did this update come out? I was only aware of "the new market wizards".

But may i propose that these guys, that said, "yeah, you can interview me because I AM a market wizard" obviously have one of those things that we all know will eventually bring about a downfall. They all have an ego.

Maybe the unsung market participants who thought, no that's not me, I aint no wizard are still going strong?

 

[donaldduke]

Hey DionysusToast,
You are pretty good at raising interesting threads! I take my hat off to you.
When did this update come out? I was only aware of "the new market wizards".

But may i propose that these guys, that said, "yeah, you can interview me because I AM a market wizard" obviously have one of those things that we all know will eventually bring about a downfall. They all have an ego.

Maybe the unsung market participants who thought, no that's not me, I aint no wizard are still going strong?

There has been no written update, not for the first two books anyway.

Some one may have done a followup article/interview in the press or on the internet, but Schwager has never published anything as far as i know.

 

[timsk]

Hi DT,
I accept that it's a possibility and, as D70 says, it's an interesting topic for a thread. And yes, you're right, in that I have a vested interest in believing that luck isn't the prime reason for successful traders making it big. If I was convinced - or become convinced - that your theory is 100% correct, I would give up trading now. However, for your theory to be right, it would mean that there is no skill involved in trading at all and that it is purely a game of chance like Roulette in which the house eventually wins. Indeed, as we all know, the longer one plays a game of chance, any lucky gains made early on will be lost in the fulness of time. The mathematical edge that favours the house always plays out eventually. My belief, or perception - call it what you will, is that pro' traders like the Wizards are not gamblers and that the markets are not a purely a game of chance. If one takes your theory to its logical extreme, then the extraordinary results of all human endeavor are all just statistical anomalies that can be attributed to lady luck. And if that was true, and everyone knew it was true, then very few of us would bother to get out of bed in the morning.

Tim.

 

[DionysusToast]

There is no such follow up interview book that i know of. Do you have a link?

Not for the first two books anyway, there may have been one for his third book which i dont own and i think contained interviews with stock traders near the height of the internet boom.

Indeed - it's the 2003 Paperback Edition.

Note that I should have said "Stock Market Wizards" which is a different tome to "Market Wizards", still a book in the series though.

Most of the original interviews took place from mid 1999 to early 2000. The follow up interviews in this version of the book took place in 2003.

http://www.infibeam.com/Books/info/...rds-Interviews-with-America-s/0066620597.html

 

[Rossini]

''I find it interesting that some people can't digest the fact that sheer mass of people in the industry combined with the laws of probability means that traders like this will exist through luck alone - even for 20 years.''

Look at it from another angle. If you owned a system that constantly won 50% of the time and had losers half the size of your winners then you would make a lot of money. Would you attribute your success to? The fact that you where lucky in designing such a system? No. You would know the reason for your success is because your system had an edge, simple as that. The fact that 20 years later the system no longer works is inconsequential. It did work and this was not down to luck it was simply a trading edge being exploited in the form of systematic trading.

 

[DionysusToast]

However, for your theory to be right, it would mean that there is no skill involved in trading at all and that it is purely a game of chance like Roulette in which the house eventually wins.

Actually - my post is more about seeing if people can accept the possibility that some or all of the people that Schwager interviewed were lucky. I expected it to be an emotive issue.

I figured most would find the idea abhorrent. Mostly for the reasons you outlined.

Now - I fundamentally disagree with what you say above. If you accept that say 50% or even 100% of Schwagers interviewees were there through luck, it does NOT in any way prove or disprove that there is no skill involved in training.

For one - the sample size is too small. Let's say that the 13 traders in the book were all lucky. You cannot extrapolate from there and say that all succesful traders are lucky.

If someone doesn't want to accept that those 13 guys may have been lucky, it could be because they falsely think this means that all succesful traders are lucky.

I would not make that assumption because it is taking a small sample size and applying it universally.

Perhaps not the best trait for a trader.

Let's say I put on a red tie tomorrow and that day I had 10 winning trades. Then a week later I put on a red tie and had another 10 winning trades. Would I be right to presume that the tie is responsible for the winning trades.

Similarly - the traders in Schwagers book have very plausible reasons for their success - but although they are more complex than the red tie, they could be equally false.

 

[timsk]

Hi DT,
Okay, I see where I've gone wrong. I made what appears to be a false assumption that you were just using the Market Wizards to hang your theory on. I've then extrapolated it and applied it (not unreasonably IMO!) to all traders across the board. So, in answer to your original point about the wizards making it big through pure luck, I'm very happy to accept that as a theory, it's entirely possible. However, if you accept (and perhaps you don't) that trading isn't a game of pure chance and that - broadly speaking - the wizards are a dedicated, talented, intelligent, clever and well informed bunch - then it's pretty unlikely that they made their fortunes through luck alone. If you were asserting that their outlier trades in which they made their biggest profits were attributable more to luck than skill and judgement, then I suspect a lot more people would readily agree with you, me among them.
Tim.

 

[zupcon]

Would it be fair then to say that a number of the wizards in these books are not wizards after all - they are just a statistical anomaly ?

I would say with 100% certainty that one or two may well be a statistical anomaly:

I've only known or met a handful of traders, but the the most profitable amongst them was extremely successful on exactly that very basis.

I'm focussing increasingly on automated strategies, and a key part of my strategy is to compare results the results from various elements of my systems against results achieved from a purely random approach. The two key issues to note are that:

a) A purely random approach (even based on day trading) can occassionally resut in realistic looking long term equity curves. If Ray the random robot can do it, then a market wizard can do it too.

b) I'm increasingly trading the same strategies diversified across different instruments, timeframes, and system parameters. Whats interesting is that the smallest of changes between those systems can lead to large differences in short term performance (and hence the need for diversification) These differences, based on the approach I'm using, are due to LUCK. Although markets are predictable (and easily predictable over the short term) anything can happen anytime, and it does.

The majority of methods used by traders (particularly in regards to entry) are no better than random chance, and the few approaches that do have an edge, have generally been inverted, and traders use the edge against themselves. For the majority of people luck plays a major role (and not just in trading), but you'll struggle to find anyone who'll admit it.

Traders have a tendancy to fool themselves they have control, hence the use of indicators for entry. The majority are hardly likely to want to confront the truth. In my limited experience, the ones who do understand the probabilistic nature of the problem that theyre dealing with, and understand that its about managing the uncertainty and risk, tend to do OK.