Discretionary trading

[manucastle]

I will close my TNA position.



I have added an IWM Robot Comment. I will probably close my TNA position today in order to avoid a possible down-gap on Monday that would be linked to an un-friendly political decision on which I have no control.

Trading in the back-ground of a big political decision is more like playing chess than a probability based investment decision.



Pascal
================================================== ====

Pascal,

As I have often stated I am extremely pleased with the VIT service and the very valuable feedback from yourself and others.

The only small problem I have is recently Billy mentioned that you and he would not publish your discretionary trades. I understand this because my worst faults over 21 years of investing have been discretionary emotional trading and overiding my systems.

Now we have the Robot this negates this major problem in my trading.

Your comment above is probably very wise and considered but is still discretionary and has me, and I am sure others, confused has to know what to do because I have a lot of respect for your opinion.

Europe, together with the Greek situation is obviously very serious and could get a lot worse over the weekend into next week.

I do not need to be led by the hand but I would appreciate a clear route - are you and Billy going to issue statements on your discretionary trading or not ?

Sorry to be so forward but this is an important topic for us all going forward and needs to be black and white IMHO.

Thanks in advance :O)

Trev

[Billy]

I will close my TNA position.



I have added an IWM Robot Comment. I will probably close my TNA position today in order to avoid a possible down-gap on Monday that would be linked to an un-friendly political decision on which I have no control.

Trading in the back-ground of a big political decision is more like playing chess than a probability based investment decision.



Pascal
================================================== ====

Pascal,

As I have often stated I am extremely pleased with the VIT service and the very valuable feedback from yourself and others.

The only small problem I have is recently Billy mentioned that you and he would not publish your discretionary trades. I understand this because my worst faults over 21 years of investing have been discretionary emotional trading and overiding my systems.

Now we have the Robot this negates this major problem in my trading.

Your comment above is probably very wise and considered but is still discretionary and has me, and I am sure others, confused has to know what to do because I have a lot of respect for your opinion.

Europe, together with the Greek situation is obviously very serious and could get a lot worse over the weekend into next week.

I do not need to be led by the hand but I would appreciate a clear route - are you and Billy going to issue statements on your discretionary trading or not ?

Sorry to be so forward but this is an important topic for us all going forward and needs to be black and white IMHO.

Thanks in advance :O)

Trev

Trev,

Thank you for the frank discussion. I am the only one who commited to stop disclosing my discretionary decisions in this forum.
Pascal never committed to such thing.
For the time being, I stick to the penny with the robot on triple leverage.
That's because my discretionary decisions since the robots are online have totalled a significant net loss compared to the robot results.
My stop for TNA is at the robot stop.
Billy

[Pascal]

I will close my TNA position.



I have added an IWM Robot Comment. I will probably close my TNA position today in order to avoid a possible down-gap on Monday that would be linked to an un-friendly political decision on which I have no control.

Trading in the back-ground of a big political decision is more like playing chess than a probability based investment decision.



Pascal
================================================== ====

Pascal,

As I have often stated I am extremely pleased with the VIT service and the very valuable feedback from yourself and others.

The only small problem I have is recently Billy mentioned that you and he would not publish your discretionary trades. I understand this because my worst faults over 21 years of investing have been discretionary emotional trading and overiding my systems.

Now we have the Robot this negates this major problem in my trading.

Your comment above is probably very wise and considered but is still discretionary and has me, and I am sure others, confused has to know what to do because I have a lot of respect for your opinion.

Europe, together with the Greek situation is obviously very serious and could get a lot worse over the weekend into next week.

I do not need to be led by the hand but I would appreciate a clear route - are you and Billy going to issue statements on your discretionary trading or not ?

Sorry to be so forward but this is an important topic for us all going forward and needs to be black and white IMHO.

Thanks in advance :O)

Trev

You are right Trev. The best is that I do not disclose what i do with my robot positions.
From now on, the Robot comments will only be technically oriented to explain what the robot is doing and why.

I will limit my "opinions" to the "comment of the day" section.



Pascal

[manucastle]

You are right Trev. The best is that I do not disclose what i do with my robot positions.
From now on, the Robot comments will only be technically oriented to explain what the robot is doing and why.

I will limit my "opinions" to the "comment of the day" section.



Pascal

Thanks very much Pascal.

Trev

[davidallison@gmail.com]

Billy or Pascal,

For a fixed amount dedicated to robot trades, would you split this amount equally between the robots, and stick to that allocation, trading all primary signals? Alternatively and diversification benefits aside would you rotate out of positions that indicate Strong primary entries for a given robot, in an effort to seek better returns, for the robot account.

Dave

[Billy]

Billy or Pascal,

For a fixed amount dedicated to robot trades, would you split this amount equally between the robots, and stick to that allocation, trading all primary signals? Alternatively and diversification benefits aside would you rotate out of positions that indicate Strong primary entries for a given robot, in an effort to seek better returns, for the robot account.

Dave

Dave, we are currently researching these optimal allocations scenarios. Pascal is doing intensive backtesting on this and he'll come soon with his conclusions.
In the meantime, there is no way to be sure that any choice is superior to another.
Billy

[Pascal]

Billy or Pascal,

For a fixed amount dedicated to robot trades, would you split this amount equally between the robots, and stick to that allocation, trading all primary signals? Alternatively and diversification benefits aside would you rotate out of positions that indicate Strong primary entries for a given robot, in an effort to seek better returns, for the robot account.

Dave

Hello Dave,


This is a very good question.
As Billy wrote it, I am now busy with the research, whihc could still last a few days, as more results bring more questions.

As of now, the basic findings are the following:
1. If you trade one instrument with high leverage (for example TNA/TZA) and do not mind the drawdown risk, then adding GDX will not help create better returns. In fact, GDX would even hurt the TNA/TZA returns.
2. If you use TWM/RWM (double leveraged), then a combination with GDX improves the returns.
3. If you use IWM and GDX then staying with only GDX is better. But then GDX is more volatile than IWM, which means larger drawdowns.

The principle of this test was to invest 100% of the capital in the first robot that issues a signal that is not neutral (Buy, strong buy, sell, strong sell). You then keep the trade unless one of the following event occurs:
- You hit a stop loss or you have a signal change
- The other robot issues a signal that is not neutral. In such a case, you sell 50% of the first position and invest 50% in the second robot.

The idea is to be 100% invested whenever there is a non neutral signal in one of the robot and to split between robots when the two robots have produced a signal.

I did not test the idea of switching 100% to the second robot from the first if the second robot signal is very strong, while the first turned to neutral. I believe that this could encourage over trading and would deny some benefits of trading two instruments that are lightly correlated. However, this is just an opinion. Testing might reveal that this opinion is groundless.

To be honest, what I like best in the case number two (TWM/RWM and GDX) is that the equity curve is very very smooth. I will not post it now, because I need to check some things and then do more tests, but as I am much a "risk averse" person, that rings a bell somewhere in my trading guts.


Pascal

[Billy]

As of now, the basic findings are the following:
1. If you trade one instrument with high leverage (for example TNA/TZA) and do not mind the drawdown risk, then adding GDX will not help create better returns. In fact, GDX would even hurt the TNA/TZA returns.
2. If you use TWM/RWM (double leveraged), then a combination with GDX improves the returns.
3. If you use IWM and GDX then staying with only GDX is better. But then GDX is more volatile than IWM, which means larger drawdowns.


Pascal,
Do you have any significant indication yet about which allocation is providing the best long term risk-adjusted returns between 100% of the time only in TNA/TZA (option 1) and a combination of TWM/RWM + GDX (option2)?
Billy

[Pascal]

Pascal,
Do you have any significant indication yet about which allocation is providing the best long term risk-adjusted returns between 100% of the time only in TNA/TZA (option 1) and a combination of TWM/RWM + GDX (option2)?
Billy

Not yet, but when you have an instrument with such strong returns as TNA/TZA, switching 50% out of that instrument into an instrument that produces only half of the return cannot help the performance (drawdowns/risks put aside). Also, the definition of a good risk/reward balance is different for each investor.

So, as I wrote, new research results bring more questions, which lead to more research and so on.
I plan to wrap what I did up to now, write a report and then do more work on the pending issues.


Pascal

[Maxime A.]

Yet another option may be to try achieving better returns w/ the GDX robot by using the leveraged ETFs NUGT and DUST. That said, GDX is already quite jumpy, so the result may be too volatile for most investors. Also, they haven't been around much, so any backtest is going to be hard. Thanks,

Max

[Rembert]

Just my oppinion but I think it is best to initially take leveraged ETF's out of the equation when searching for the optimal use for a given amount of capital.
Otherwise the test will probably indicate it's best go with TNA/TZA only for maximum gain.

The problem I see with that is that this is based on historical data. The maximum drawdown of any system lies in the future and is unknown.
Risk of ruin is probably quite high going all in on TNA/TZA even tough historical tests show it maximizes profit.

Instead of trying to maximize gain it might be better to maximze the MAR ratio. (MAR = CAGR / Max DD) This way, risk is taken into account.
My guess is diversification in systems leads to a higher MAR ratio and to a smoother equity curve. I believe that it is safer to use leverage on a diverse portfolio of systems (either by margin or leveraged ETF's) than it is to do so on a single system, no matter how great historical backtest says that system is.

[grems8544]

Just my oppinion but I think it is best to initially take leveraged ETF's out of the equation when searching for the optimal use for a given amount of capital.
Otherwise the test will probably indicate it's best go with TNA/TZA only for maximum gain.

The problem I see with that is that this is based on historical data. The maximum drawdown of any system lies in the future and is unknown.
Risk of ruin is probably quite high going all in on TNA/TZA even tough historical tests show it maximizes profit.

Instead of trying to maximize gain it might be better to maximze the MAR ratio. (MAR = CAGR / Max DD) This way, risk is taken into account.
My guess is diversification in systems leads to a higher MAR ratio and to a smoother equity curve. I believe that it is safer to use leverage on a diverse portfolio of systems (either by margin or leveraged ETF's) than it is to do so on a single system, no matter how great historical backtest says that system is.

Completely concur with Ray's comment. If I may add some color ...

Risk adjusted positioning increases the work we must put into our trading activities. It basically throws out the concept of "I'll own a maximum of 10 positions and my portfolio is Y, so each position is Y/10". While this approach certainly works, it is far less than optimum as the number of positions gets less and less. As M&K, O'Neil, et al. advise, fewer positions is best. If you believe this, then the impact of MDD is crucial as the drawdown contains more of your capital. Unfortunately, most individuals who counsel a low number of holdings rarely address risk-adjusted position sizing, because it's not a Finance 101 concept.

The obvious question is how to correctly position size. The not-so-obvious answer is dependent upon:

1) the expected gain of the securities in the entire portfolio
2) the correlation of the securities in the entire portfolio
3) the variance of the individual securities in the entire portfolio
4) the standard deviation of the prices of the individual securities
5) the allocation of capital between the different components of the individual securities

A very good starting point on understanding these concepts can be found in the following PowerPoint presentation. I've zipped it because ".ppt" is not an available upload option.

8905

The key concept here, which directly supports Ray's assertions, are that we start the optimization process by constructing a portfolio with a minimization of variance in mind. Here, variance is equated to drawdown. Hence, for a given set of equities with quantified gains as well as standard deviation of the price series, we can form a portfolio which reduces variance while increasing gain.

The PowerPoint shows how to do it for a two-security portfolio -- doing it for more (which obviously is practical) requires more work and further requires that you know how to use Excel. The outline of how to do this for N-portfolios is also in the PowerPoint, and relies heavily on the work of Simon Benninga (http://www.amazon.com/Financial-Modeling-Simon-Benninga/dp/0262026287/ref=sr_1_1?ie=UTF8&s=books&qid=1308581862&sr=8-1). The book has a disk with macros and worksheets ready to go, but of course, you have to understand some of the basic concepts shown in the PowerPoint in order to get anything meaningful out of the book.

Further supporting Ray's assertion that we should not use leveraged instruments in our optimization process is this paper:

8906

This is a heavier read, but pay attention to Figures 1 and 2, which give you some framework around the "optimum" leverage levels for the general market over various periods in the past. This paper is relevant because leveraged instruments multiply BOTH volatility as well as gains, resulting in the same behavior as their non-leveraged counterparts during the optimization process (I ignore the impact of the daily rebalancing of leveraged daily ETFs, which is adverse over the longer term).

Finally, the approach in the book and the included paper above do not address in a concise manner that volatility changes with the time frame being measured; they use standard lookback periods and take the volatility as equal-weighted over the lookback period. This is not a good approach in real life, so this last paper is a (heavy) introduction on using a moving average concept to have a better volatility estimate:

8907

Riskmetrics pioneered research on risk management, and I've not found a better reference.

If you've made it this far and you have a basic understanding of the concepts of volatility, jump to chapter 5, specifically Table 5.7 onward, to get a view on how the exponentially weighted moving average (EWMA) model is used in the forecast of returns and variances.

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

Disclaimer: This is all a work in progress, and I'm still learning, learning, learning. Portfolio/position sizing is HARD, which is why most people (including me) simply do the Y/10 per position and are done with it. The concept of minimal-variance portfolio construction is hard to achieve, yet maintain in a practical fashion, but this shouldn't make you throw your hands up and say "it's not worth it". It's clear it *is* worth it, especially if you are happy with your holdings and have the knowledge that you're closer to maximum efficiency in gains and reduction in drawdown than you'd ever be with a Y/10 approach.

[Billy]

Paul,

This is a terrific wealth of resources on the topic! It shows your commitment to search for constant improvement.
Before digging into all the details, I have a conceptual hesitation.
It seems all the models are based on individual securities or stocks.
Since IWM is already including a +/- 2,000 stocks diversification and GDX is including a 32 stocks diversification, shouldn't you take into account the exisiting diversification implied in ETF holdings? In other words, can we be sure that the models are not misleading for our robot allocation needs?
Billy

[Rembert]

Thanks for the material Paul. This is a topic that interests me vey much. It goes beyond simply being a trader. It's about learning how to manage money ... large sums of money. No doubt that all the billionaire funds know these concepts all too well.

Billy, I think one can diversify in multiple ways. Using IWM is a great starting point, being diversified in 2000 different stocks. But it doesn't have to end there. One can add another layer of diversifiction by employing multiple systems using different markets/strategies/timeframes. It doesn't have to be 1000 systems, I think 3 (IWM,GDX,XLE) is a nice number that is still managable.

Here's a few more links from the excellent automated trading system blog :

http://www.automated-trading-system.com/trading-diversification-free-lunch/ (I like the coffee cup analogy)
http://www.automated-trading-system.com/trick-reduce-drawdowns/

[asomani]

I would consider to:

1. start by constructing a portfolio of the IWM Robot and GDX Robot that has produced the lowest intraday drawdown, not considering leverage.

2. next construct a portfolio of the IWM Robot and GDX Robot that has produced the lowest end-of-day drawdown, not considering leverage.

3. next construct a portfolio of the IWM Robot and GDX Robot that produces the lowest annualized volatility, not considering leverage.

After performing the above steps, a clear answer as to the appropriate allocation between the two robots may emerge (at least based on historical results) for the maximization of volatility-adjusted return. If not, further investigation will be required.

[Pascal]

I would consider to:

1. start by constructing a portfolio of the IWM Robot and GDX Robot that has produced the lowest intraday drawdown, not considering leverage.

2. next construct a portfolio of the IWM Robot and GDX Robot that has produced the lowest end-of-day drawdown, not considering leverage.

3. next construct a portfolio of the IWM Robot and GDX Robot that produces the lowest annualized volatility, not considering leverage.

After performing the above steps, a clear answer as to the appropriate allocation between the two robots may emerge (at least based on historical results) for the maximization of volatility-adjusted return. If not, further investigation will be required.

This looks like a great idea. I accept your help with great pleasure, because I am somewhat overwhelmed.
In the attached file, you will have two sets of thee colums. These are the results of either IWM/GDX or TWM/GDX combined portfolio (compared to each element separately).

I'd be happy if you could share the formula that you use to calculate the drawdowns that you mention on these equity curves. I can then generate different ratio of IWM/GDX combinations as I only use now 50/50.

Thanks again for your help.



Pascal

8914

[asomani]

This looks like a great idea. I accept your help with great pleasure, because I am somewhat overwhelmed.
In the attached file, you will have two sets of thee colums. These are the results of either IWM/GDX or TWM/GDX combined portfolio (compared to each element separately).

I'd be happy if you could share the formula that you use to calculate the drawdowns that you mention on these equity curves. I can then generate different ratio of IWM/GDX combinations as I only use now 50/50.

Thanks again for your help.



Pascal

8914

Hi Pascal,

Actually, I wasn't offering to help at this time. I would love to help but, you see, I'm quite busy right now with my work and other activities. I simply wanted to suggest a few ideas to you and the group to help get things rolling a bit.

I have taken a bit of time, though, to make an example for you of how to calculate drawdowns in Excel. I have attached a spreadsheet showing how the calculations should be performed, based on my knowledge. In my spreadsheet, the EOD (end-of-day) and intraday drawdown calculations are for a simple strategy that takes a buy-and-hold approach on the SPY from the beginning of the year through to yesterday.

I hope the spreadsheet is clear. Please let me know if you have questions.

You will notice that, to calculate intraday drawdown, you will need to obtain the intraday % change for the Robots based on the High of the day (if short) or Low of the day (if long) for each day of the testing period.

[Pascal]

Hi Pascal,

Actually, I wasn't offering to help at this time. I would love to help but, you see, I'm quite busy right now with my work and other activities. I simply wanted to suggest a few ideas to you and the group to help get things rolling a bit.

I have taken a bit of time, though, to make an example for you of how to calculate drawdowns in Excel. I have attached a spreadsheet showing how the calculations should be performed, based on my knowledge. In my spreadsheet, the EOD (end-of-day) and intraday drawdown calculations are for a simple strategy that takes a buy-and-hold approach on the SPY from the beginning of the year through to yesterday.

I hope the spreadsheet is clear. Please let me know if you have questions.

You will notice that, to calculate intraday drawdown, you will need to obtain the intraday % change for the Robots based on the High of the day (if short) or Low of the day (if long) for each day of the testing period.

Thanks. I'll have a look.


Pascal

[Rembert]

Amibroker is probably more suited to test such things. But since you're more used to work with Excel that's probably the better option.

It's possible to import the signals for each system into Amibroker and generate the equity curves.
After that it should be quite easy to test various things.

Amibroker also has optimization ...
http://www.amibroker.com/guide/gifs/h_optimize.gif
The X and Y axis are the parameters to opimize for example position size etc. Z axis is the value you opimize for (CAGR, Max DD, MAR, ...).

Here's an example I've been looking at regarding testing a portfolio of different systems. Haven't had the time yet to test it out tough.

http://quantingdutchman.wordpress.com/2011/02/16/simulating-a-portfolio-of-strategies-in-amibroker-12/
http://quantingdutchman.wordpress.com/2011/04/18/simulating-a-portfolio-of-strategies-in-amibroker-22/

[Rembert]

It's nice to know intraday DD, but not really necessary I think.
Again this article makes a good case for testing closed equity DD instead of intraday DD.
http://www.automated-trading-system.com/trick-reduce-drawdowns/

[grems8544]

Since IWM is already including a +/- 2,000 stocks diversification and GDX is including a 32 stocks diversification, shouldn't you take into account the exisiting diversification implied in ETF holdings? In other words, can we be sure that the models are not misleading for our robot allocation needs?

Sorry for the delay in posting. I've been away since the subsequent conversation ensued.

Billy, I wrestle with this concept often. Is holding one security at 100% in my portfolio, such as the IWM or the SPY, "diversified", or does proper diversification require multiple holdings? I think this is what you are asking. I don't know the answer as I bang this out but allow me to think through it as I write...

When we talk about diversification, indirectly I think we're talking about a reduction in risk. Ideally, we're also looking for maintaining/increasing potential returns while reducing risk, but at the end of the day, I think the goal is to minimize risk while maintaining some constant (predictable) level of potential gains. I add this latter part because the "risk free return" (RFR) reference used in many calculations is a 3 month T-bill, which is assumed to have 0 volatility at some nominal return.

"Diversification Benefit" is something that the financial types talk about when going from a portfolio containing n=1 security to a portfolio containing n=m (m=many ~>15, 20, 50, etc.). The "benefit" portion comes from a reduction in risk as we add stocks. Offsetting this benefit are transactional costs, e.g., it's far cheaper to manage 1 security than to mange 50. So, returns are enhanced as n ---> 1, but risk increases as n ---> 1.

IWM is the market, for all intensive purposes. In support of this statement here is the correlation of IWM with other markets:

8947

The correlation with the Dow is 0.87, with the S&P500 0.92, and the NASDAQ 0.95. Being the "market", we can assume that it is diversified, e.g., addition of any further stocks to IWM will not result in a significant change in the correlation with other broad markets. Put another way, we can assume that the addition of another security to IWM will not change the volatility of IWM in a significant manner.

Correspondingly, holding a single position that is 100% in IWM certainly represents a portfolio that contains the market. For the portfolio to be diversified, we need to successfully show that additional positions in the portfolio do not reduce volatility and return. Put another way, for a given level of estimated return, can we find a security that has greater or equivalent estimated return but lower volatility?

First of all, it's REALLY hard to find a company, or companies, that fit this bill. With my processing power, I can process about 55 stocks at a time, and I went through over 600 stocks in 10 batches and didn't find one that had higher return at the same or less volatility than IWM. Here's a representative chart:

8942

I've circled IWM -- you can see that there are no stocks with higher return and lower volatility -- so it's the combination of all of these stocks, with their relative correlations, that give IWM higher return but lower volatility.

Take a look at this next figure:

8943

Look at the left -- you can see IWM, and you can see a purple diamond just below IWM. The blue line represents all the possible combinations of these 54 stocks in terms of weighting, and the one "closest" to the IWM is chosen. The purple diamond shows that this portfolio of 54 stocks can be made to look like IWM with specific weightings, which is what managers of big funds attempt to accomplish.

Hence, I think we can conclude that:

1) Holding 100% IWM is to be considered fully diversified within your portfolio, e.g., you're not going to improve your Diversification Benefit from n = 1 (because transaction costs are minimized, which maximizes takehome gain), and
2) out of the universe of 1800+ small-cap stocks, there are NO stocks that give an expected return higher than IWM at an equivalent or lower volatility (although I only checked ~600 of them).


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

IWM was the hard nut, GDX is easier.

Answering the question of whether a portfolio holding 100% GDX is diversified portfolio follows along the same lines as above. If we can show that the addition of a security improves/maintains the portfolio return while reducing volatility, then we can say GDX is not a diversified holding.

Take a look at the following chart, which is the expected returns/volatility of the components of GDX, sans LIHR:

8944

I've circled GDX to make it stand out on the chart. As you can see, there are INDEXES (^RUT, ^IXIC, and ^GSPC) that have lower volatility at equivalent or higher gain. Furthermore, because ^RUT has the highest expected return but is lower in volatility, then the IWM proxy should be considered as the additional security so that we can maximize return yet realize a lower volatility than GDX alone.

From the chart above we can conclude that owning a portfolio that was comprised of 100% GDX is not diversified -- there are additional securities that can be added which reduce risk (volatility) while improving gain.


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

Extra Credit ( I'm on a roll now and am learning a few things ... )

The next question in my mind is "what allocation of IWM and GDX minimizes volatility?". This is the same as asking what combination minimizes my pain to drawdown?

Take a look at this chart:

8945

The unlabeled yellow diamond in the middle of the picture is the result of 100 shares being allocated to GDX, and 100 shares being allocated to IWM. The blue line should extend down there but for some reason it does not. I'll look into that later.

It's clear from the blue line that there is yet another combination that has higher gain but minimum volatility. That allocation ends up being:

73% IWM
27% GDX

This is telling me that all things being equal, using the price data that has a ~ 100d moving average applied (see RiskMetrics document provided earlier), that a minimum-risk portfolio exists with IWM and GDX in the allocations above.

Of course, maximizing expected return means placing all your eggs in IWM, since it has the highest return yet the lower volatility of the two, or IWM = 100% and GDX = 0%.

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

Extra Extra Credit ...

Since Pascal is working on XLE, let's see what the allocations for a IWM-GDX-XLE triplet should be in order to minimize risk in the portfolio, again using data through 6/20/11:

8946

So, we see that XLE lies halfway between IWM and GDX in terms of volatility as well as expected return. The unlabled yellow diamond is a 100-share contribution in each security. The end of the blue line just above the unlabled diamond is the place where risk (volatility) is minimized -- and this corresponds to an allocation of

IWM: 69%
GDX: 23%
XLE: 8%

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

So, in summary,


trading 100% in IWM maximizes your expected return and can be considered a fully diversified holding
trading 100% in GDX reduces your potential gain as well as introduces additional volatility to your portfolio
trading 73% in IWM and 27% in GDX minimizes your portfolio volatility, at the cost of some expected return


I hope this has been helpful -- I've learned a few things, especially that intuition isn't always correct (my assumptions about GDX standing alone).

[Kenneth K]

So, in summary,


trading 100% in IWM maximizes your expected return and can be considered a fully diversified holding
trading 100% in GDX reduces your potential gain as well as introduces additional volatility to your portfolio
trading 73% in IWM and 27% in GDX minimizes your portfolio volatility, at the cost of some expected return




That may be accurate if you replace 'trading' with the word 'holding'. Holding the different portfolios over time, you may end up with such risk-return characteristics. But when it comes to 'trading', none of that matters. None of those parameters come into material effect during the very short trading horizon of the Robots. Short term trading yields portfolios with totally altered risk-return characteristics.

What is efficient for a buy and hold index portfolio, is not necessary true for an actively traded portfolio. You cannot compare apple and oranges, eventhough they are both fruits and delicious.


Trying to pinpoint an optimal mix of Robot based trading in IWM/GLD is an exercise in futility. There will be periods when one specific mix is optimal. And that optimal mix is optimal for that particular period, in direct relation to the market movement in IWM and GLD during that particular time. And naturally, that will simply fluctuate over time.


I see a lot of effort is expended on slicing and dicing the Robot signals to get the most out of it. From my limited experience in the business, I would venture to guess that there is inherent risk in over optimizing within a very small data set. The nature of the market is change, eventhough very often, and ironically, the more things change, the more they stay the same. There is an inherent risk that what works perfectly for the last three years, may not work as well in the next three years. The key is the robustness of the algorithm.

It appears that the ROBOT methology of combining a trend following component with a mean reversion component is a very robust method that will withstood the test of different market conditions, with minor retuning over time. If people over optimize it for a very narrow trading period, the result may be counter productive.


As always, best of luck to everyone.

[adam ali]

Paul,

This is very good research. As we all know, correlations can change mightily during times of market stress, and usually not for the good (i.e., they rise) right when you need their benefit. It would be interesting to see how correlations worked during such times.

[grems8544]

@Adam: completely agree that correlations move with time.

A good side experiment would be for me to check these IWM/GDX correlations on the weekend, and see if they move much. I'll post the results here in the Algo forum and we can keep track; I'll start a new thread. My intuition tells me that they won't, but now that I've written those words, they will most likely change dramatically and all bets are off :)

I should have also stated in the note that the snapshots provided in the figures are just that -- snapshots. Expected returns will change daily/weekly/whatever, as will volatility. Time frame for lookback is very important, but with the EWMA method I use, I'm smoothing a lookback over about 100 days and since it's exponential, the data between 100 days ago and 93 days ago isn't that relevant compared to all of the other data...

Time will tell...

[Rembert]

Paul,

Interesting reseach. That said, I agree with Kenneth in that the suggested IWM/GDX allocations are usefull only when one uses 'buy and hold' of these securities. But that's not the case with the robots.

To get the real picture one needs to do these 'buy and hold' tests on the equity curves of the robots instead of the ETF's directly. I'd be very interested in the results of such a test. But this is something only Pascal and Billy can do as it would require the actual historical trades of the robots. Actually thinking about it, we don't need the historical trades. Pascal could supply the equity curve data instead.

It's true that there is a limited 'data set' but that's not necessarily a problem as this post shows with 'walk forward allocation'.
One could use a sliding window of past 3 years of data to keep the allocation % dynamically up to date.
http://www.automated-trading-system.com/practical-leverage-space-model-more-realistic/

PS, anyone know the answer to the 'summer puzzle' ?
http://www.automated-trading-system.com/off-the-grid/

[Billy]

To get the real picture one needs to do these 'buy and hold' tests on the equity curves of the robots instead of the ETF's directly.
I'd be very interested in the results of such a test. But this is something only Pascal and Billy can do as it would require the actual historical trades of the robots.


Thank yoy Paul for the extraordinary depth of your research.
Thank you Kenneth and Rembert for your very constructive remarks.

In all friendliness, let me say that it is very easy to say "what to do", but it is another thing to get everything done!

Pascal is working very hard on this, but a new software development is required first for complete results.
The study is progressing on trading only the 20 DMF signals without the risk management from the robots.
Some drawdowns look frightful so the added value from the robots' risk management element will be instructive also once the new software development is finished.
Billy

[Rembert]

Hi Billy,

That's very true. But also remember we're just having a conversation here. Exchange of idea's etc.
I'm not asking Pascal or you to drop what you are doing to do some tests. No need to feel pressure from the forums to do things.

In due time I think it would be good tough to have a look into position allocation tests for the robots. But it's not a priority. When the time arrives I'll be glad to help if you guys can then provide eiter historical trades or equity curve data of the robots.

Again, really appreciate the work you guys are doing.

[Pascal]

That may be accurate if you replace 'trading' with the word 'holding'. Holding the different portfolios over time, you may end up with such risk-return characteristics. But when it comes to 'trading', none of that matters. None of those parameters come into material effect during the very short trading horizon of the Robots. Short term trading yields portfolios with totally altered risk-return characteristics.

What is efficient for a buy and hold index portfolio, is not necessary true for an actively traded portfolio. You cannot compare apple and oranges, eventhough they are both fruits and delicious.


Trying to pinpoint an optimal mix of Robot based trading in IWM/GLD is an exercise in futility. There will be periods when one specific mix is optimal. And that optimal mix is optimal for that particular period, in direct relation to the market movement in IWM and GLD during that particular time. And naturally, that will simply fluctuate over time.


I see a lot of effort is expended on slicing and dicing the Robot signals to get the most out of it. From my limited experience in the business, I would venture to guess that there is inherent risk in over optimizing within a very small data set. The nature of the market is change, eventhough very often, and ironically, the more things change, the more they stay the same. There is an inherent risk that what works perfectly for the last three years, may not work as well in the next three years. The key is the robustness of the algorithm.

It appears that the ROBOT methology of combining a trend following component with a mean reversion component is a very robust method that will withstood the test of different market conditions, with minor retuning over time. If people over optimize it for a very narrow trading period, the result may be counter productive.


As always, best of luck to everyone.

I agree with you Kenneth.

What Paul's excellent work implies is that IWM B/H makes sure that you move with everybody else and at the same speed. This means that if everybody else jumps off the cliff, you are sure to be with them. This means that the implied drawdowns for both IWM and GDX are pretty bad in buy and hold situations (shown below). This means that market diversification will not prevent bankrupcy when the whole market sinks.

The only way to avoid that is to start trading them. Then, the implied IWM/GDX balance will have more to do about the type of trading strategy we use on each instrument and on their combination.

For example, the IWM/GDX combination would be different when trading the robots than when trading a 50/200MA crossover strategy.



Pascal

8950

8951

[Pascal]

Below is a link to the IWM/GDX portfolio analysis.


http://www.effectivevolume.eu/content/Reports/IWM_GDX_Portfolio_Analysis.pdf



Pascal

[grems8544]

Below is a link to the IWM/GDX portfolio analysis.
http://www.effectivevolume.eu/content/Reports/IWM_GDX_Portfolio_Analysis.pdfl

Thank you Pascal. Great documentation, as always.

With an IWM historical MDD of -10.5%, is it safe to conclude that using the leveraged triple ETF would result in a MDD in excess of -31.5%?

Regards,

pgd

[grems8544]

What is efficient for a buy and hold index portfolio, is not necessary true for an actively traded portfolio.

Kenneth,

I agree with you on this point.

The concept of bounding is implicit in my posting; my lookback was of ~100d or so. If the results here become the boundary of expectations in the present market climate, and if further risk management (entry at 1/3 RR, stops based on pivots, etc.) gives us the suggestion that we will be bounded by the "worse-case" results presented (e.g., risk management improves performance and/or reduces risk), then I think we can have a reasonable expectation of outperforming the boundary going forward.

My conclusions on allocation were simply to suggest that there is an allocation which represents, in the present market climate, a configuration that reduces volatility of an entire portfolio that is comprised of two trading instruments. I think we can agree that if holding 100% IWM, we have a reasonable expectation that the volatility will be lower than holding 100% GDX. If the last 100d of actual returns and volatility represents the next few days/weeks (forward looking) of expected returns and expected volatility, then I think my conclusions are not too far from the mark, independent of whether you are holding IWM, GDX, or a mixture of both.

Kind regards,

pgd

[Pascal]

Thank you Pascal. Great documentation, as always.

With an IWM historical MDD of -10.5%, is it safe to conclude that using the leveraged triple ETF would result in a MDD in excess of -31.5%?

Regards,

pgd

If you use leverage blindly, this will be the case, but knowing that there is an astounding 0.49 correlation factor between the MDD and the 20DATR, what you need to do is to avoid tripple leveraged when the ATR is above a certain level (my hunch is above 3%) and avoid double ETF when the ATR is for example above 5%. This way, you really reduce the MDD, but you keep your potential for big gains. I would need more research to find out the real combination though.


Pascal

[Rembert]

Thank you for the analysis Pascal. Seems like Pareto's law applies to the robot signals as well in that 80% of the profits come from 20% of the signals. One of great things about the VIT robot is that it's able to tell up front which signals are likely to fall into that 20% category. This is something most systems can only do in hindsight.