Moving Average as an indicator for entry into the market?

[CanOz]

CanOz,

I would recommend doing some research on dividend yield of the ASX20 compared to a "risk free return" such as a 1 year term deposit at a leading bank. When there is a large spread between the dividend yield and the risk free return, it is time to enter the market using an income based fund. When there is little or no spread between the dividend yield and the risk free return you want enter the market using a growth based fund. Or is the other way round?

Dividend yield is the fundamental key.

Cheers

odds-on

Thanks Oddson, i see your thesis...basically when fixed interest provides no benefit, then in theory the equity market should have good yields and capital growth potential, like an indicator of the stage in the business cycle yeah?

CanOz

 

[johenmo]

We are talking about long term fundamental investing? You can only do fixed fractional position sizing if you adopt a price based stop. IMO if I bought a share for a specific fundamental reason, I would not have a price based stop.

I used to think that way. Now I have a stop loss in nearly every situation - suits me, and saves me from myself. I believe it's a good idea for a beginner. They can then follow it after that to see if they saved money or missed out on it.

 

[odds-on]

Thanks Oddson, i see your thesis...basically when fixed interest provides no benefit, then in theory the equity market should have good yields and capital growth potential, like an indicator of the stage in the business cycle yeah?

CanOz

I see it as a crude indicator of investor sentiment. The key is the spread compared to long term average spread. My research on the web indicates that it is a reasonable indicator, obviously DYOR. The contrarian in me thinks that an investor should buy the income based fund when the spread is low and the growth based fund when the spread is high - it would be interesting to do some backtesting over a few years.

Cheers

Oddson

 

[sinner]

I see it as a crude indicator of investor sentiment. The key is the spread compared to long term average spread. My research on the web indicates that it is a reasonable indicator, obviously DYOR. The contrarian in me thinks that an investor should buy the income based fund when the spread is low and the growth based fund when the spread is high - it would be interesting to do some backtesting over a few years.

Cheers

Oddson

I've done some testing on this already.

Short version:
Assume you split the ASX20 50:50 based on yield percentiles and end up with two baskets "growth" (low yield) and "income" (high yield). You can split the index into as many smaller baskets as you want and the phenomenon still is present e.g. using the 3 highest yield vs 3 lowest yield instead of 10 vs 10...

Your "sentiment indicator", along with similar measures like the dispersion of the ratio between growth/income baskets and ratio momentum, will provide better performance than the market benchmark.

However, this doesn't mean you aren't exposed to market price risk! Rather that you will lose a few % less than the market when it goes down and gain a few % more than the market when it goes up (largely due to beta factors though!!!). Your equity curve still largely resembles the index.

I've also done some testing on "market neutral" versions i.e. long growth/short income or short growth/long income depending on dispersion and these can outperform regardless of underlying market conditions, however due to market beta factors, these models seem better suited for long and short volatility trades (ratio chart resembles the VIX) than fundamental exposure.

 

[odds-on]

I've done some testing on this already.

Short version:
Assume you split the ASX20 50:50 based on yield percentiles and end up with two baskets "growth" (low yield) and "income" (high yield). You can split the index into as many smaller baskets as you want and the phenomenon still is present e.g. using the 3 highest yield vs 3 lowest yield instead of 10 vs 10...

Your "sentiment indicator", along with similar measures like the dispersion of the ratio between growth/income baskets and ratio momentum, will provide better performance than the market benchmark.

However, this doesn't mean you aren't exposed to market price risk! Rather that you will lose a few % less than the market when it goes down and gain a few % more than the market when it goes up (largely due to beta factors though!!!). Your equity curve still largely resembles the index.

I've also done some testing on "market neutral" versions i.e. long growth/short income or short growth/long income depending on dispersion and these can outperform regardless of underlying market conditions, however due to market beta factors, these models seem better suited for long and short volatility trades (ratio chart resembles the VIX) than fundamental exposure.
View attachment 47577

Thanks Sinner,

Perhaps outperformance could be achieved by purchasing XSO based fund when the ASX20 dividend yield spread is high compared to 1 year term deposit rates. When the spread returns to average purchase a fixed income fund. Alternate capital between the two funds based on the dividend yield spread of the ASX20 and 1 year term deposits.

I do not have any experiencing applying these types of systems, just ideas i got from the web. I guess there maybe an edge of a couple of percentage.

Cheers

Oddson

 

[joea]

I recently purchased "Unholy Grails" by Nick Ridge.
Although I have not completely read it all, It has a strategy section with simulations.
MA is used in conjunction with strategies as a filter(XAOA) All Ordinaries Accumulation Index.
I have added the post here because of the MA implications.
Interesting book. Different to most.
Simulations by AmiBroker, with Tradestation to confirm signals and performance.
Well done!!

Book endorses the principle:::

"Absorb what is useful,
Discard what is not,
Add what is uniquely your own".
Bruce Lee

My comment(after reading 20% and glancing through the book), it will help make me a better trader.
It is one of the books I have read that does not go over the same old stuff about (the complete trading scenario).

But I will always give credit to any book, because there is always something that you can pick up.
It is how you compile these bits together that is the challenge. I am sure this book will shows that.
joea

 

[CanOz]

Thanks Sinner,

Perhaps outperformance could be achieved by purchasing XSO based fund when the ASX20 dividend yield spread is high compared to 1 year term deposit rates. When the spread returns to average purchase a fixed income fund. Alternate capital between the two funds based on the dividend yield spread of the ASX20 and 1 year term deposits.

I do not have any experiencing applying these types of systems, just ideas i got from the web. I guess there maybe an edge of a couple of percentage.

Cheers

Oddson

Thanks for the discussion fellas...as you can see this fits my desire for some kind of a switch...

Joel, if I could manage our own super then we would employ the 20%flipper to either Australian or US equities as we do with our long term growth fund currently. It is because we cannot control the equities we purchase, only the sectors as listed on Mercers investment choices that I have thought of an alternative approach. Great book!

CanOz

 

[sinner]

Thanks for the discussion fellas...as you can see this fits my desire for some kind of a switch...

Joel, if I could manage our own super then we would employ the 20%flipper to either Australian or US equities as we do with our long term growth fund currently. It is because we cannot control the equities we purchase, only the sectors as listed on Mercers investment choices that I have thought of an alternative approach. Great book!

CanOz

Personally if that was the case for me I'd investigate "sector momentum", it's a well established phenomenon with better underlying rationale than single name momentum (i.e. if you account for sector momentum and/or futures momentum where applicable, single name momentum doesn't necessarily outperform).

Here's a well known example
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1585517

 

[johenmo]

Personally if that was the case for me I'd investigate "sector momentum", it's a well established phenomenon with better underlying rationale than single name momentum (i.e. if you account for sector momentum and/or futures momentum where applicable, single name momentum doesn't necessarily outperform).

Here's a well known example
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1585517

Agreed... I found that looking at the sector a stock is in and how that is moving relative to the rest of the market adds a few points in my favour. Sectors have different levels of popularity (sentiment) at any particular moment in time.

CanOz - this would help with the flipper strategy, I think.

 

[tech/a]

Agreed... I found that looking at the sector a stock is in and how that is moving relative to the rest of the market adds a few points in my favour. Sectors have different levels of popularity (sentiment) at any particular moment in time.

CanOz - this would help with the flipper strategy, I think.

Has anyone done any walk forward testing on this.
I'm afraid my investigation into this logical hypothesis
Couldn't get me an edge.

Looking back on history in any number of charts it's pretty clear that there " should " be an edge
It's clear you can see it.

But for a sector to out perform it needs constituents that out perform not only the primary index but
Those in it's sector.
I found that often it was just one stock which hauled the sector higher.
By the time the sector showed signs of interest or out performance the stock ( at best a couple of stocks) had well and truly run their race.Often their moves were spectacular but by the time you found them and the index on the move there was little left going forward.

What is needed and I never found it was a way of anticipating the move rather than observing past history.

 

[Ves]

The Australian market is probably too small for a sector by sector based strategy. Most sector weightings are dominated by a few big caps. It might work better in overseas markets.

 

[tech/a]

The Australian market is probably too small for a sector by sector based strategy. Most sector weightings are dominated by a few big caps. It might work better in overseas markets.

Don't think it's about size.
It's about application.
Putting the theory into practice
Profitably. That's why I questioned
Forward testing. Easy to see in hind site
Don know about putting it into practice.

 

[sinner]

I'm afraid my investigation into this logical hypothesis
Couldn't get me an edge.

No offense but if you couldn't find an edge I'd be more inclined to ask what was your investigation based on exactly? If you read the paper I posted, and testing doesn't confirm it's results, I will eat my freaking hat. How exactly are you trading "past winners"?

By the time the sector showed signs of interest or out performance the stock ( at best a couple of stocks) had well and truly run their race.Often their moves were spectacular but by the time you found them and the index on the move there was little left going forward.

just lols for this, as I look at the XUJ chart from Aug 2011 compared to XEJ. What you are saying here shows me you don't understand the momentum phenomenon as described in financial literature. I don't want to continue this discussion until you've read at least the paper I posted above.

Oh look, it only took me a few second to pull a bunch of work on this already:
http://marketsci.wordpress.com/2010/05/18/roundup-fundztrader-sector-rotation-strategy/
http://ibankcoin.com/woodshedderblog/2010/01/20/fidelity-select-sector-rotation-strategy-wrap-up/
http://engineering-returns.com/2011...ive-sector-rotation-model-to-improve-returns/

 

[Uncle Festivus]

In my share trading days I used to use a deviation from moving average to catch the tops or bottoms of a trend for a reversal - more for trading than investing? Get some charting software that you can create your own indicators eg Bullcharts.....

 

[tech/a]

Sinner

How on earth do you read that I don't understand sector momentum.

I looked at comparison of the All Ords v the various sectors.
Sure sectors out performed the comparison and within the
Sector stocks out performed the index --- dragging it above
The comparison.

Of course you can see it.
Have you FORWARD TESTED IT!
Or point me to any papers that have?

Run a walk forward test here
You'll soon see what I mean
Or perhaps we'll see what you mean

 

[Trembling Hand]

Has anyone done any walk forward testing on this.
I'm afraid my investigation into this logical hypothesis
Couldn't get me an edge.

Thats a very interesting comment. How much money, especially retail money, chases "logical hypothesis". I reckon my "career" has been based on fading logical setups. The reason is twofold. One there is heaps of volume at the logical setups like breakouts. Two the first to spew up positions are those that act on "logical hypothesis" because they are surprised when they get taken off-side.

Maybe we should set up a thread where some can put up some "logical hypothesis" and others can do a walk forward test. I would be very interested in the results. Must be a walk forward though - no back testing.

 

[sinner]

Sinner

How on earth do you read that I don't understand sector momentum.

Specifically because of your statement:

By the time the sector showed signs of interest or out performance the stock ( at best a couple of stocks) had well and truly run their race.

I think you are mixing up trend following and momentum.

I looked at comparison of the All Ords v the various sectors.
Sure sectors out performed the comparison and within the
Sector stocks out performed the index --- dragging it above
The comparison.

"looked at"? How, exactly, did you "look at" the comparison? I don't understand what you're trying to say here, other than the phenomenon is present?

Of course you can see it.
Have you FORWARD TESTED IT!
Or point me to any papers that have?

Run a walk forward test here
You'll soon see what I mean

Or perhaps we'll see what you mean

Yes, I have. Have you? I know the return profiles of momentum trading strategies and its alpha versus the index in different volatility regimes. I know when to leverage momentum and when I'd be best off just holding the underlying. Do you? Or are you just "preaching fact" based on what? There are a bunch of mutual funds in the US that trade pure equity momentum strategies and a few more that trade risk adjusted momentum, surely you agree these constitute a walk forward test? I also trade my own strategy on single names, with an index filter very similar to the one you use for techtrader. So yes, I have walk forward traded.

If you actually care (which I doubt), check out Kenneth French website, as he actually tracks all this sort of data, walk-forward, every month.

Considering the attitude I got on the last thread I tried to trade a strategy (Bill Williams breakouts) live on this forum, no way do I plan on attempting to share in a new thread, I classify that as a waste of time now, would rather spend that time actually trading.

Thats a very interesting comment. How much money, especially retail money, chases "logical hypothesis". I reckon my "career" has been based on fading logical setups. The reason is twofold. One there is heaps of volume at the logical setups like breakouts. Two the first to spew up positions are those that act on "logical hypothesis" because they are surprised when they get taken off-side.

First of all
http://www.iijournals.com/doi/abs/10.3905/jii.2011.2.3.050

This article fills this gap, finding that ETFs representing sectors experiencing positive (negative) momentum have higher (lower) returns on days associated with the execution of this momentum strategy. It also finds that calendar days coinciding with the implementation of this rule are associated with increased trading volume in related sector ETFs, particularly on the buy side in more recent periods.

Second, it's impossible to get 'taken off-side' and spew up a position on a momentum strategy, unless you aren't actually following the strategy.

EDIT: I will at least concur some of the more robust momentum strategies include a lag to account for short term (<1 month) mean reversion.

...can't believe I'm trying to convince you guys the only phenomenon that all the literature agrees exists in pretty much all markets, exists!

 

[Trembling Hand]

it's impossible to get 'taken off-side' and spew up a position on a momentum strategy, unless you aren't actually following the strategy.

Huh? So they have a 100% hit rate?

 

[Joules MM1]

Huh? So they have a 100% hit rate?

i'm keen to read the answer to that, too

 

[sinner]

Huh? So they have a 100% hit rate?

No I don't mean that, I only meant that in a momentum strategy you don't spew your position 'cos it got taken offside (like you might if you traded a MA crossover), you are only taking exits on portfolio rebalance and time exits. There is no performance basis exit.