Why aren't there more long term investors?

[nobel_1101]

I'd suggest broadening the scope of your consideration to include the possibility that a long-term trader might be in more than one position at a time (and not simply in stocks), as well as the impact of reinvestment on drawdowns. It makes all the difference.

As an example, I've been long futures for metals, grains, energies, and several softs (e.g. cotton, coffee) for some time now. Each of these markets has experienced significant, sharp retracements nearly simultaneously:
Gold -18%
Silver -24%
Crude Oil -16%
Chicago Wheat -29%
MPLS Wheat -42%
Cotton -25%
Sugar -24%
Cocoa -25%
Coffee -27%

These are leveraged instruments making large percentage moves against my positions, yet somehow my portfolio has experienced only a 2.4% drawdown. A similar market event occurred last July/August, with a similar small impact on my portfolio. Either miracles are taking place or maybe, just maybe, there is something to what I'm saying here.

FWIW, a fund's offering memorandum or disclosure document typically lays out all but the most intimate details of their trading methodology. It's actually relatively easy to ascertain the style used by a manager. Most advertise it openly. This allows a straightforward comparison of risk-adjusted returns across styles for anyone sufficiently motivated to perform one.

jj

Ok you seem to have answered my question (albeit in a roundabout way). I use stocks purely for illustration of this discussion.

So if all your positions declined almost simultaneously an average of -25% and your account only reduced by 2.4% then fair enough. But what's not possible (and this goes back to qwertyuiop1 original question) is for this long-term system to make the same large profits as a short term system.
From your own figures, if all your positions increased by 100% then your account will only increase by 9.6%.
You can't have it both ways, either the long term system has low drawdowns and low profits, or high drawdowns and high profits.

 

[nine]

I think you are making assumptions that are incorrect nobel but lets see what Mathemagician has to say about that. I suspect that "nearly simultaneously" may be less so than we would expect - either that or other components of the portfolio moved ahead strongly at the same time. Thank you for getting things back on a civil basis.

 

[Mathemagician]

I think you are making assumptions that are incorrect nobel but lets see what Mathemagician has to say about that. Thank you for getting things back on a civil basis.

This is correct, and it is critical to always strive to not confuse assumptions with conclusions. Just because you've assumed it to not be possible doesn't mean it's not. It only means you'll never be able to do it until the self-limiting assumption is abandoned.

The real key to understanding the performance of any trading program is the ratio of return to risk. This is because one can use leverage to produce either low risk/low return or high risk/high returh, but the ratio stays constant. This particular program has produced a raw compound annual growth rate of 34.8% (in reality).

Note that an increase in position size of 100% would approximately double the drawdown, not quadruple it. As you probably already know, the addition of leverage has a nonlinear impact on the results and this impact is path-dependent, with the departure from linear increasing with aggressiveness.

Sorry if this is a bit rambling, but it's getting late.

jj

P.S. No, that was not a complete list of my positions. It's just a recent example of a time when several of the markets in a long-term portfolio retraced significantly (this was not even a complete list of the markets that retraced significantly) but the portfolio as a whole did not.

 

[nobel_1101]

This is correct, and it is critical to always strive to not confuse assumptions with conclusions. Just because you've assumed it to not be possible doesn't mean it's not. It only means you'll never be able to do it until the self-limiting assumption is abandoned.

The real key to understanding the performance of any trading program is the ratio of return to risk. This is because one can use leverage to produce either low risk/low return or high risk/high returh, but the ratio stays constant. This particular program has produced a raw compound annual growth rate of 34.8% (in reality).

Note that an increase in position size of 100% would approximately double the drawdown, not quadruple it. As you probably already know, the addition of leverage has a nonlinear impact on the results and this impact is path-dependent, with the departure from linear increasing with aggressiveness.

Sorry if this is a bit rambling, but it's getting late.

jj

P.S. No, that was not a complete list of my positions. It's just a recent example of a time when several of the markets in a long-term portfolio retraced significantly (this was not even a complete list of the markets that retraced significantly) but the portfolio as a whole did not.

I'm simply trying to extract a logical argument from you as to why you think its possible to trade long term positions (lasting many months to a few years) with low drawdown but yet achieve the same return possible from a short term system.

I was refering to an increase in price of 100% not size of position. In any case, mathematically it just doesn't add up. Either you have one of three things:
1. Have low drawdowns and low potential profits
2. Have high drawdowns and high potential profits
3. Are scaling in/out of positions when the market is favourable or adverse respectively. By definition this is not a true long term system but more a long term/short term hybrid.

Happy trading

 

[Splitlink]

Oh, mp, you make it sound so easy :whistling. It's a theory I've been trying to prove to my longer term "active investor" wife getting on for 35 years. Unfortunately, she outdoes me most years and remains to be convinced Alright, I know, I must be a lousy trader, but you'd have thought I'd have got the hang of that simple "be long during long periods and short during short" by now .

good trading

jon

As long as you are not on wife support, that's the main thing. You're not a lousy trader, Jon, unless it's her bank account over there.

 

[Mathemagician]

I'm simply trying to extract a logical argument from you as to why you think its possible to trade long term positions (lasting many months to a few years) with low drawdown but yet achieve the same return possible from a short term system.

I was refering to an increase in price of 100% not size of position. In any case, mathematically it just doesn't add up. Either you have one of three things:
1. Have low drawdowns and low potential profits
2. Have high drawdowns and high potential profits
3. Are scaling in/out of positions when the market is favourable or adverse respectively. By definition this is not a true long term system but more a long term/short term hybrid.

Happy trading

I surrender. Good luck!

OK, I can't resist one more try...

1 and 2 are the same thing! The only difference is leverage. Here's why...

Suppose you've got two trading programs, Program A and Program B.

Program A has a an average annual return of 10% and a maximum drawdown of 5%.
Program B has a an average annual return of 50% and a maximum drawdown of 50%.

What makes program A better than program B? Well, program A has a return/drawdown ratio of 2.0, while program B has a return/drawdown ratio of 1.0. This means A has a better risk/reward ratio than B.

Now, suppose as an investor you want a 50% annual return target. You could either choose program B at 1x leverage or program A at 5x leverage.

Suppose instead you are an investor who wants a 10% annual return target. You could either choose program B at 0.2x leverage or program A at 1x leverage.

The risk/reward ratio captures the relative performance between the two programs, and leverage determines the absolute returns. Does that make sense?

Given that daytrading and long-term trading have exhibited similar risk/reward ratios over long periods of time, the only thing that can get you a higher absolute return out of one vs the other is leverage and leverage is available to both programs.

The fact that long-term traders usually do not trade with very high leverage is a reflection of their risk preferences and not an inherent limitation of the strategy. I would attribute this to the fact that long-term diversified trading requires larger accounts than daytrading, and thus long-term traders tend to have a lower appetite for risk. This also means that daytrading appeals to traders with small accounts who are trying to run them up into big accounts quickly. The ability to operate in a small account is the true advantage of daytrading, while capacity is the symmetric advantage of long-term trading. The optimum solution is a mixture of both.

jj

 

[Mathemagician]

I think you are making assumptions that are incorrect nobel but lets see what Mathemagician has to say about that. I suspect that "nearly simultaneously" may be less so than we would expect - either that or other components of the portfolio moved ahead strongly at the same time.

Ah, this wasn't there the first time I looked!

These moves were just about simultaneous, and quite recent in fact. A properly constructed portfolio can help absorb the impact of such events, and as you've deduced these are only a handful of positions among a total of roughly 80. Of course the other positions are correlated to these positions, some more and some less, but overall there is sufficient diversification available across the global futures markets to absorb most events like these.

jj

 

[nobel_1101]

I surrender. Good luck!

OK, I can't resist one more try...

1 and 2 are the same thing! The only difference is leverage. Here's why...


The fact that long-term traders usually do not trade with very high leverage is a reflection of their risk preferences and not an inherent limitation of the strategy. I would attribute this to the fact that long-term diversified trading requires larger accounts than daytrading, and thus long-term traders tend to have a lower appetite for risk. This also means that daytrading appeals to traders with small accounts who are trying to run them up into big accounts quickly. The ability to operate in a small account is the true advantage of daytrading, while capacity is the symmetric advantage of long-term trading. The optimum solution is a mixture of both.

jj

1. and 2. are not the same. If you traded using 1. you would still be able to trade 10 years from now. If you traded using 2. you would have been wiped out after year 3 at best.

Suppose you've got two trading programs, Program A and Program B.

Program A has a an average annual return of 10% and a maximum drawdown of 5%.
Program B has a an average annual return of 50% and a maximum drawdown of 50%.

What makes program A better than program B? Well, program A has a return/drawdown ratio of 2.0, while program B has a return/drawdown ratio of 1.0. This means A has a better risk/reward ratio than B.

Now, suppose as an investor you want a 50% annual return target. You could either choose program B at 1x leverage or program A at 5x leverage.

Suppose instead you are an investor who wants a 10% annual return target. You could either choose program B at 0.2x leverage or program A at 1x leverage.

The risk/reward ratio captures the relative performance between the two programs, and leverage determines the absolute returns. Does that make sense?

Yes that does make sense but what does that have to do with this discussion?

Given that daytrading and long-term trading have exhibited similar risk/reward ratios over long periods of time, the only thing that can get you a higher absolute return out of one vs the other is leverage and leverage is available to both programs.

The fact that long-term traders usually do not trade with very high leverage is a reflection of their risk preferences and not an inherent limitation of the strategy. I would attribute this to the fact that long-term diversified trading requires larger accounts than daytrading

You've just conflicted yourself here. You suggest that the long term trade can have the same drawdown as a short term trade and have the same profit potential as a short term trade. If that was the case (which is isn't) how it is that long-term trading requires larger accounts?

The risks (measured in absolute price terms) between a long term and short term trade are not the same. The long term trade is much higher. Think about it. A longer term trade would keep a wider stop. Yes?
So by definition the long trade is risking more. And to compensate, the position size is reduced so that the amount of capital at risk is on par with the short term trade. But by doing so you're reducing the amount of profit you can make on the long-term trade (because your position size is lower). Does that make sense?

I hope that's clear. I do take your point about diversifying your portfolio by trading different systems of different time horizons, so long as the benefit gained from the reduction in trade correlation outweighs the increase in opportunity cost by missing out on more short term trades.

 

[Mathemagician]

Unless you are willing to move beyond a "one trade only" mindset, there is no real point in continuing. It's plainly obvious that a single short-term trade will tend to have a nearer exit in absolute terms than a long term trade. I can't imagine anyone would dispute that. However, this says nothing about the risk/return profile over time of a portfolio of short-term trades relative to a portfolio of long-term trades.

Just let me know if you're interested. If not, no harm done, no hard feelings, and enjoy the rest of your day!

jj

P.S. Wish I could throw a few more pips at you just for being a reasonable chap!

 

[nobel_1101]

Unless you are willing to move beyond a "one trade only" mindset, there is no real point in continuing. It's plainly obvious that a single short-term trade will tend to have a nearer exit in absolute terms than a long term trade. I can't imagine anyone would dispute that. However, this says nothing about the risk/return profile over time of a portfolio of short-term trades relative to a portfolio of long-term trades.

Just let me know if you're interested. If not, no harm done, no hard feelings, and enjoy the rest of your day!

jj

P.S. Wish I could throw a few more pips at you just for being a reasonable chap!

Thanks. I'm glad we've finally cleared that up.
Yes I'm interested to know how you view the situation to be different if considering multiple positions.

 

[mp6140]

MP -- wish I could disagree with you Jon !

Oh, mp, you make it sound so easy :whistling. It's a theory I've been trying to prove to my longer term "active investor" wife getting on for 35 years. Unfortunately, she outdoes me most years and remains to be convinced Alright, I know, I must be a lousy trader, but you'd have thought I'd have got the hang of that simple "be long during long periods and short during short" by now .

good trading

jon

======================================================

jon --- in my role as dutch uncle Im forced often to state the truth and that may well be that youre not really good at locating the points where the market reverses, although a simple ma crossover system will do that, and ive seen plenty around (and they all seem to work except in a ranging situation)

therefore, Im afraid I must agree with your statement --- if you try using the LRC's coupled with a simple RSI, I would imagine you to do a whole lot better, as long as you go flat whenever the price hits the LRC top or bottom ---- i got newbs making money where they never did before, which is why my "reputation" points do so well in such a short time, and coupled with my sl policy, seems to work for them.

furthermore, what i teach is free ---- Im a trader and not a businessman selling potato chips --- If you make more money, youre not taking any away from me, so WHY shouldnt i share ??

enjoy and trade well

mp

 

[Mathemagician]

Cool.

When considering multiple positions one is kind of forced to move towards marked-to-market periodic return series instead of trade-by-trade return series. This is because inconsistencies can arise otherwise. As an example, when scaling into and out of a position with just 2 lots (2 market-systems) one might buy at 100, buy at 110, sell at 105, and then sell at 115. This "trade" can be grouped in the following ways:

1. One trade with an average buy of 105 and an average sell of 110.
2. Two winning trades: buy 100 / sell 105 and buy 110 / sell 115.
3. One winning trade and one losing trade: buy 100 / sell 115 and buy 110 / sell 105

The choice made is arbitrary and many of the statistics people compute are different depending on which choice is made (e.g. drawdown, percent profitable, profit factor, win/loss ratio, etc.). (Note that some unscrupulous system vendors will take advantage of this to exaggerate their claims, as it is relatively easy to hide massive intra-trade drawdowns and create artificially high win rates just by manipulating trade-by-trade analysis.) That means that as soon as one introduces a second market-system, one can change one's analysis based on choices that they arbitrarily make. This is obviously a problem and must be addressed

One highly effective solution to this is to forego trade-by-trade analysis and analyze the daily/weekly/monthly marked to market returns, which are independent of any choices made by the researcher (so long as the underlying is sufficiently liquid to allow price discovery, which is required for honest marked-to-market analysis and the lack thereof for certain derivatives is manipulated by many a bank/fund to their advantage and their clients' disadvantage).

It is on this level, daily/weekly/monthly period returns, that long-term and short-term portfolios exhibit similar risk-adjusted returns over time when examined carefully. Note further that a similar relationship tends to hold true for systematic and discretionary traders, despite what gurus would have you believe, though that is a topic for another day and another time.

Your thoughts?

jj

 

[nobel_1101]

When considering multiple positions one is kind of forced to move towards marked-to-market periodic return series instead of trade-by-trade return series. This is because inconsistencies can arise otherwise. As an example, when scaling into and out of a position with just 2 lots (2 market-systems) one might buy at 100, buy at 110, sell at 105, and then sell at 115. This "trade" can be grouped in the following ways:

1. One trade with an average buy of 105 and an average sell of 110.
2. Two winning trades: buy 100 / sell 105 and buy 110 / sell 115.
3. One winning trade and one losing trade: buy 100 / sell 115 and buy 110 / sell 105

The choice made is arbitrary and many of the statistics people compute are different depending on which choice is made

Hmm...People who fiddle the results of their trades in this way are begging to be wiped out! There's only one real way of measuring the performance and that's by keeping track of which system made which trade. i.e. system 1 bought at 100 and sold at 105, system 2 bought at 110 and sold at 115, say. Anything else just isn't reality.

(e.g. drawdown, percent profitable, profit factor, win/loss ratio, etc.). (Note that some unscrupulous system vendors will take advantage of this to exaggerate their claims, as it is relatively easy to hide massive intra-trade drawdowns and create artificially high win rates just by manipulating trade-by-trade analysis.) That means that as soon as one introduces a second market-system, one can change one's analysis based on choices that they arbitrarily make. This is obviously a problem and must be addressed

One highly effective solution to this is to forego trade-by-trade analysis and analyze the daily/weekly/monthly marked to market returns, which are independent of any choices made by the researcher (so long as the underlying is sufficiently liquid to allow price discovery, which is required for honest marked-to-market analysis and the lack thereof for certain derivatives is manipulated by many a bank/fund to their advantage and their clients' disadvantage).

It is on this level, daily/weekly/monthly period returns, that long-term and short-term portfolios exhibit similar risk-adjusted returns over time when examined carefully. Note further that a similar relationship tends to hold true for systematic and discretionary traders, despite what gurus would have you believe, though that is a topic for another day and another time.

Your thoughts?

jj

I'm sorry, you're still not giving me logical arguments, only waffle based on vague 'experiences'. Its surely not that hard to break down into concise points if what you're saying truely has merit (and I would be willing to listen if you could do that).
But I've run out of time (and patience) on this discussion.

Happy trading

 

[Mathemagician]

Wow. Yeah, it's probably best that we stop here.

jj

 

[nine]

OK. I didn't find a simple answer to one question in the thread from this point. So I will ask a couple of questions:

1. Why was the drawdown of your portfolio only 2.4%?
2. What is the portfolio's likely annual return (likely can be defined as you wish)?

Just point me at the answer if its already been given.

I'd suggest broadening the scope of your consideration to include the possibility that a long-term trader might be in more than one position at a time (and not simply in stocks), as well as the impact of reinvestment on drawdowns. It makes all the difference.

As an example, I've been long futures for metals, grains, energies, and several softs (e.g. cotton, coffee) for some time now. Each of these markets has experienced significant, sharp retracements nearly simultaneously:
Gold -18%
Silver -24%
Crude Oil -16%
Chicago Wheat -29%
MPLS Wheat -42%
Cotton -25%
Sugar -24%
Cocoa -25%
Coffee -27%

These are leveraged instruments making large percentage moves against my positions, yet somehow my portfolio has experienced only a 2.4% drawdown. A similar market event occurred last July/August, with a similar small impact on my portfolio. Either miracles are taking place or maybe, just maybe, there is something to what I'm saying here.

FWIW, a fund's offering memorandum or disclosure document typically lays out all but the most intimate details of their trading methodology. It's actually relatively easy to ascertain the style used by a manager. Most advertise it openly. This allows a straightforward comparison of risk-adjusted returns across styles for anyone sufficiently motivated to perform one.

jj

 

[Mathemagician]

OK. I didn't find a simple answer to one question in the thread from this point. So I will ask a couple of questions:

1. Why was the drawdown of your portfolio only 2.4%?
2. What is the portfolio's likely annual return (likely can be defined as you wish)?

Just point me at the answer if its already been given.

Sorry about that...

1. Just lucky, I guess. Had some other positions zig while these zagged, and they were weighted in such a way as to prevent the adverse moves from dominating the portfolio.
2. I'd guess about a 20% expected return at this leverage, though I tend to focus much more on risk expectations than return expectations. Returns kind of just happen if the risk is taken care of.

jj

 

[stockstrader]

I'm sorry, you're still not giving me logical arguments, only waffle based on vague 'experiences'.

We are clearly reading different posts. jj's responses have been far from illogical and "waffle."

 

[nine]

It is true that jj's posts have been far from waffle. I suspect (could be wrong) that nobel was seeking a couple of direct answers that jj may not have been aware were needed. So in that context, discussions of the surrounding area were, perhaps, waffly. Hence my two questions above (which I am pleased jj answered to my satisfaction).


The overall issue of day trading vs position trading seems to attract misunderstandings on both sides. And you get people arguing past each other as they miss each others points.

It is absolutely true that short term trading has greater potential than long term because the markets are fractal and one can profit from the sub-fractals in the big one ... Thus, in theory one can risk the same percentage on every trade and, if the long and short term strategies had the same expectancy make an X% profit much more frequently. Yes, one can argue different expectancies but the fact is that the REALLY big profits you see on the P&L thread on another board come from day traders. And its for only one reason - they can risk X% of their accounts for a Y% expected return 5-100 times per day.

Problems occur though. The simplest is the scaling one where short term strategies can (often) only scale up so far before slippage etc stop them working ... and this is what jj seems to have focussed on. More complex is an old broker's observation that the more often an account holder trades the faster they seem to go out of the business - IMO this is because as you go shorter term you get more transaction costs, more slippage, less decision time (lower quality) and more gambling behaviours. So the promise of shorter term trading is often simply not realised.

That is why I recommend people learn to trade on a swing/position timeframe and only move to day trading if that is working for them.

 

[luckyd1976]

That is why I recommend people learn to trade on a swing/position timeframe and only move to day trading if that is working for them.

Very good advice... Start out LT or MidT swing trading to get a feel for the market Price Actions. Also, it takes time to learn your own personal trading style/psychology.

IMHO diversification into different time frames is more important to me than the asset classes.

 

[Mathemagician]

It is true that jj's posts have been far from waffle. I suspect (could be wrong) that nobel was seeking a couple of direct answers that jj may not have been aware were needed. So in that context, discussions of the surrounding area were, perhaps, waffly. Hence my two questions above (which I am pleased jj answered to my satisfaction).


The overall issue of day trading vs position trading seems to attract misunderstandings on both sides. And you get people arguing past each other as they miss each others points.

It is absolutely true that short term trading has greater potential than long term because the markets are fractal and one can profit from the sub-fractals in the big one ... Thus, in theory one can risk the same percentage on every trade and, if the long and short term strategies had the same expectancy make an X% profit much more frequently. Yes, one can argue different expectancies but the fact is that the REALLY big profits you see on the P&L thread on another board come from day traders. And its for only one reason - they can risk X% of their accounts for a Y% expected return 5-100 times per day.

Problems occur though. The simplest is the scaling one where short term strategies can (often) only scale up so far before slippage etc stop them working ... and this is what jj seems to have focussed on. More complex is an old broker's observation that the more often an account holder trades the faster they seem to go out of the business - IMO this is because as you go shorter term you get more transaction costs, more slippage, less decision time (lower quality) and more gambling behaviours. So the promise of shorter term trading is often simply not realised.

That is why I recommend people learn to trade on a swing/position timeframe and only move to day trading if that is working for them.

Again, I am not saying that one is better than the other. (I actually do both types of trading for myself and for clients.) My position is that each has its strengths and weaknesses, that there is a place and purpose for each type of trading, and that neither style is inherently superior to the other. In fact, the optimum scenario would be having the ability to create a balanced portfolio of uncorrelated trading approaches, two components of which would be long-term diversified and daytrading. This is what multistrat funds attempt to do and if they do it well they beat any individual style hands down.

jj