Dynamic asset allocation strategy through composite leading indicator

[yiugn]

So many times I am confused whether to implement dynamic asset allocation strategy or just simply adhering to 1/N allocation.

I have tried to develop trustworthy market timing signal to use dynamic asset allocation by trend following strategy, but only to fail, not because I could not generate beautiful equity curve, but

it's fundamental problem, overoptimization and curve-fitting, lack of universality.

Alternative approach I tried was risk parity or volatility based approach which I found was more reliable and logical. I used 10month trailing volatility based on coefficient of variation, which was found to be

quite successful, although it's controversal to implement this approach considering tax problem. However, at least it was successful in my backtest using stock and bond. At least, it has shown superior less

volatility compared to simple 1/N approach over unstable market condition.

I think this approach should be differentiated by conventional market timing approach, as it is proven that usually increasing volatility is related to higher risk, and usually it is assoaciated with

market decline. Risk parity or volatility based method does not expect any direction of market, but it only control risk at the same level between negatively correlated assets, generating alpha not by predicting

market direction but controlling risk.

Another approach I've recently tried which was found to be quite successful is using composite leading indicator for asset allocation.

It is well known that composite leading indicator is a very reliable fundamental indicator predicting and reflecting future movement of market price, with less whipsaw compared to market price itself.

In my approach, if composite leading indicator(monthly basis) increase on 3 consecutive period, stock/bond ratio = 75/25, and decrease 3 consecutive periods the ratio is reversed.

The allocation ratio of stock/bond depends on your personal taste. If you are aggressive it could be 80/20, and if you are conservative it could be 60/40.

I have tested this strategy which was quite impressing, but I don't know how I should post the equity curve and the result.

Yes, it's a kind of market timing strategy, on which quite a lot of you may show abhorrance, "Market timing is totally useless, garbage, trash, overoptimization....etc..."

but I think it's fundamentally different from conventional market timing strategy on these aspects.

1. Not price action based technical approach, fundamental based approach

2. By using leading indicator for market index, conventional time-lag problem for market timing or trend following strategy is somewhat compensated.

3. Less whipsaw and superior smoothness compared to market price only based approach

For this reason, although simple 3 consecutive month based momentum strategy is a kind of market timing approach, the result is completely different whether it is based on fundamentally leading indicator or

mechanical market price itself.

I recommend you to try this approach, composite leading indicator could be obtained from http://stats.oecd.org/Index.aspx?datasetcode=MEI_CLI

[stone]

yiugn, is the idea that "the market" fails to properly price in the information from these "leading indicators"? So what you are doing is exploiting a persistent mispricing that results from "the market" foolishly giving too much weight to inferior indicators or whatever?

[yiugn]

Yes, CLI has 2 month lag and it's obvious that the signal is neither 100% leading to real market price nor it always succeed to enable 'top selling' or 'bottom fishing'

Sometimes, as you know, it's signal can be completely opposite to market movement. It's quite natural!

I am never saying that this kind of approach is worth considering because it's signal is always perfect. It's totally absurd we can expect 'perfect' signal with these kind of 'market timing' or 'risk parity' or

'CLI' approach.

Although sometimes it produce false signal, quite a lot it can also profitable signal. Although it fails to succeed in 'bottom fishing' and 'top selling' quite a lot of time and has 'lagging' feature,

it has advantage of indicating timing for more allocation to trending asset, therefore producing more alpha compared to 1/N allocation.

It is not reasonable to claim this approach is useless simply only based on it cons. We should always take into account of pros/cons. at the same time.

Conventional 1/N approach is completely free from market timing or market direction, virtually no chance of losing 'bottom' or 'top', as it's always invested, that's the greatest benefit of this approach.

However, the problem of this approach is that it's relatively inefficient in trending condition, because we should always take 'both' direction at the same time. This is drawback of 1/N allocation.

Therefore, putting together pros/cons at the same time, I cannot dare to say neither 1/N allocation nor dynamic strategy is always superior than the other.

Both strategies have completely opposite characteristics, thus it's not reasonable to claim either approach is always dominant over the other. It's just different.

CLI 'does' make mispricing, it's absurd to expect any perfect signal with whatever kind of approach.

If you try simply with 'monthly price', not by 'CLI', you can get quite similar result.

[stone]

yiugn, do you think that there is any truth in the idea that markets are becoming less "trending" and more "stochastic"?

[MediumTex]

yiugn,

How long have you actually been using the strategy you are descibing and what has your actual performance been with it?

If this strategy works, why isn't everyone doing it?

I always wonder about these things when people offer new trading systems.

[clacy]

yiugn,

How long have you actually been using the strategy you are descibing and what has your actual performance been with it?

If this strategy works, why isn't everyone doing it?

I always wonder about these things when people offer new trading systems.

Why isn't everyone using the PP if it works?  The fact is there are millions of market participants and not all of them have the same goals, ideas for strategies, etc.

There is no "one and only way" to do things, IMO and there are continually new ideas that going forward, may or may not work as well as what the backtesting suggests, which includes the PP.

[MediumTex]

yiugn,

How long have you actually been using the strategy you are descibing and what has your actual performance been with it?

If this strategy works, why isn't everyone doing it?

I always wonder about these things when people offer new trading systems.

Why isn't everyone using the PP if it works?  The fact is there are millions of market participants and not all of them have the same goals, ideas for strategies, etc.

There is no "one and only way" to do things, IMO and there are continually new ideas that going forward, may or may not work as well as what the backtesting suggests, which includes the PP.

I kind of think that the OP may have been spam, and I was just fishing for a post that would help me confirm that the poster was a real person rather than some kind of post-generating machine.

To answer the question about why everyone isn't using the PP, it goes against many core beliefs about the nature of what is knowable in life.  I wouldn't anticipate that more than a tiny percentage of investors would ever use something like the PP, given the level of humility it requires and the way it requires you to always be a little out of step with most other investors.

If, however, any one trading system actually consistently worked as advertised I don't know why everyone wouldn't start using it, which would of course cause it to immediately stop working.  It would be nice if the people selling the trading systems would include this in their commercials.

[blancalily]

If, however, any one trading system actually consistently worked as advertised I don't know why everyone wouldn't start using it, which would of course cause it to immediately stop working.  It would be nice if the people selling the trading systems would include this in their commercials.

MT, Do you ever worry that you're killing the goose that lays the golden eggs by promoting the PP?

[MediumTex]

If, however, any one trading system actually consistently worked as advertised I don't know why everyone wouldn't start using it, which would of course cause it to immediately stop working.  It would be nice if the people selling the trading systems would include this in their commercials.

MT, Do you ever worry that you're killing the goose that lays the golden eggs by promoting the PP?

I don't think that Craig and I will ever reach a fraction of the people that Harry Browne reached through his books and other writings.

I just think that the PP is a niche strategy that will always make perfect sense to a few people and will seem crazy to the rest.

[clacy]

If, however, any one trading system actually consistently worked as advertised I don't know why everyone wouldn't start using it, which would of course cause it to immediately stop working.  It would be nice if the people selling the trading systems would include this in their commercials.

MT, Do you ever worry that you're killing the goose that lays the golden eggs by promoting the PP?

I don't think that Craig and I will ever reach a fraction of the people that Harry Browne reached through his books and other writings.

I just think that the PP is a niche strategy that will always make perfect sense to a few people and will seem crazy to the rest.

Agreed.  I think that the PP is just too radical for most people, even those that don't hold a lot of education in the financial arena. 

When you think about the so-called professionals, 99% of the "investment community" has been conditioned that there is basically one way to invest, which involves stock heavy portfolios.  The conventional wisdom is that during your working years you hold stocks/bonds in a range of 80/20 on the top end to 40/60 on the low end.  Then when you retire you convert most if not all your assets to annuities or municipal bonds until you die.  Most of these folks are incentivised to sell products, so the PP is counter-productive to their income and career path.

Maybe 2% of market participants have even heard of the PP, maybe 1/10 of those would ever take the plunge.

The other factor is that the PP will underperform the S&P many years, and I think there is something about underperformance periods that make them almost as difficult to handle as drawdowns.

I just can't envision a day in my lifetime where a significant portion of investors are willing to hold 25% in gold, and only 25% in stocks.

[clacy]

MT, Do you ever worry that you're killing the goose that lays the golden eggs by promoting the PP?

I don't think that Craig and I will ever reach a fraction of the people that Harry Browne reached through his books and other writings.

I just think that the PP is a niche strategy that will always make perfect sense to a few people and will seem crazy to the rest.

Agreed.  I think that the PP is just too radical for most people, even those that don't hold a lot of education in the financial arena. 

When you think about the so-called professionals, 99% of the "investment community" has been conditioned that there is basically one way to invest, which involves stock heavy portfolios.  The conventional wisdom is that during your working years you hold stocks/bonds in a range of 80/20 on the top end to 40/60 on the low end.  Then when you retire you convert most if not all your assets to annuities or municipal bonds until you die.  Most of these folks are incentivised to sell products, so the PP is counter-productive to their income and career path.

Maybe 2% of market participants have even heard of the PP, maybe 1/10 of those would ever take the plunge.

The other factor is that the PP will underperform the S&P many years, and I think there is something about underperformance periods that make them almost as difficult to handle as drawdowns.

I just can't envision a day in my lifetime where a significant portion of investors are willing to hold 25% in gold, and only 25% in stocks.

Gold is still not understood by most people and "gold bugs" are considered quacks.

Stocks are still over-respected, IMO and it would probably take another decade or more of outperformance by gold before the masses would consider switching to a PP-like portfolio.  Just about that time of course, stocks would take off and lure everyone back to a traditional portfolio.

[moda0306]

In my experience, people in "the industry" are nowhere close to where they will start using the PP.

[Alanw]

In my experience, people in "the industry" are nowhere close to where they will start using the PP.

Financial Planners and anyone in "the industry" will never use the PP because there is no money in it for them.  Where are they going to make their commissions when you only rebalance once every 1 to 3 years.

As far as the skeptics go, most will never be converted because:

Gold:  You have to be crazy to buy gold at these prices.  Gold has been in a bull market for 10 years and is due for a crash.
But whose to say gold can't be in a 20 year or even a 30 year bull market.

LTT:  Rates have nowhere to go but up.  After all bonds have been in a 30 year bull market and are due to correct.
But whose to say bonds can't be in a 40 year or 50 year bull market.  Ten years ago, bonds were in a 20 year bull market and look what happened.

The fact is, we just don't know.  That is the point many people are not willing to come to grips with.  They will go on  taking advice from their financial planner and think that his knowledge will produce favorable returns.

As for me, I think I'll go make a sandwich and get a good nights sleep.

[cowboyhat]

The truth is that most of the people currently invested in the PP, including most of the people on this forum, will quietly drop the PP in the next bull market in stocks. That is what happened last time. Anyone who is currently invested in the PP and hasn't given some serious thought to how they are going to deal with prosperity is setting themselves up to fail.

As attractive as PP gains are now, they are going to feel like losses when you compare them to returns from a 60/40 or an 80/20 portfolio in a stock bull market. If you think you are just going to hold on and accept PP under performance then it is likely (based on history) that you don't understand yourself very well.

A VP is one way to address this problem of tracking error in good times, and now is a good time to put heavy thought into how to time the stock market in a future VP. My thought is that if you are attracted to the PP, a timing method that involves lots of trades is probably not going to be something that works well for you from an investor psychology perspective. It might be better to look at the average market P/E and make a big stock buy for your VP with the intention to hold when market P/E gets down to 10 or less.

A market P/E that low seems crazy now, but the baby boomers have not yet started trying in earnest to unload the stocks they accumulated.

[stone]

cowboyhat, what will you class as the next bull market in stocks? If the FTSE100 gets above 8000, I'll be spooked rather than being overcome with exuberance.

[cowboyhat]

stone,

If you google "Plexus Asset Management" plus "PE ratio" plus "Real Returns"  you will find an article with a really interesting bar graph of the historical relationship between P/E of S&P500 stock purchases and returns over the following ten years from 1871 to 2010. I printed a copy out and have it pinned to my bulletion board at work so that I see it every day when I first sit down. Something to look at while I wait for my computer to boot up.

The messages I draw from that graph are that historically:
1) the S&P500 is a bargain when priced at PE ratio less than or equal to 10
2) the S&P500 is even odds to lose money when priced at PE ratio greater than 20 

A rational approach to buying stocks for a VP would seem to be to purchase them in a regular way when PE drops below 10, hold them between 10 and 20, and begin a program of regular selling above 20. I'm a human, however, and therefore reason is something I can only aspire toward. That's why I look at that graph everyday, trying to pound a little reason into my lizard brain. Trying to teach myself to think like Graham and Dodd, and not run around like Chicken Little when there is blood in the streets.

[MachineGhost]

A rational approach to buying stocks for a VP would seem to be to purchase them in a regular way when PE drops below 10, hold them between 10 and 20, and begin a program of regular selling above 20. I'm a human, however, and therefore reason is something I can only aspire toward. That's why I look at that graph everyday, trying to pound a little reason into my lizard brain. Trying to teach myself to think like Graham and Dodd, and not run around like Chicken Little when there is blood in the streets.

The market briefly dipped to a CAPE of 10 back in March 2009.  But that kind of V-bottom is an aberration.  Instead a long, drawn-out, being sandpapered to death is more normal, i.e. Japan.

An effective market timing approach needs to actually have a timing component; lagging fundamental indicators like P/E are more accurately described as sentiment indicators as they reflect the expansion and contraction in sentiment of what people want to pay for any given current level of earnings.

MG

[MachineGhost]

1. Not price action based technical approach, fundamental based approach

2. By using leading indicator for market index, conventional time-lag problem for market timing or trend following strategy is somewhat compensated.

3. Less whipsaw and superior smoothness compared to market price only based approach

For this reason, although simple 3 consecutive month based momentum strategy is a kind of market timing approach, the result is completely different whether it is based on fundamentally leading indicator or
mechanical market price itself.

What exactly is a "Composite Leading Indicator" and how is it composed?  What do you do with the output?

MG