(Not so) Smart Beta

[DeepState]

Any views?


Source: FT

 

[Faramir]

Hi DeepState
Can I offer any contribution? (I am really struggling to understand it - I am struggling to evaluate its merits or lack of merits from your view?)

Paragraph 1): Last sentence stating volatile markets and uncertain economic climate. This cycle has always happened. Smart Beta should have been "invented" or "used" or at least sold to the naive investor for many years. This is the first time I have heard of Smart Beta - I must be very naive.

Paragraph 3): Please help me. It shows my lack of understanding of how Fund Managers work. Especially when they need to re-balance their portfolios according to the Top ASX 50, ASX 100, ASX 300????

Paragraph 4): Alternative weighing schemes like volatility or dividends. I cannot understand how this can produce a better result. Some companies borrow to pay dividends (to artifically hold up a share price that doesn't deserve to be at the level?) I need some explanation please.

Paragraph 5): "to take advantage of perceived systematic biases or inefficencies market" OK Everyone thinks they can do this. Everyone except me since I am a beginner. The manager is "passive", so therefore it is cheaper than an "active" manager.

Paragraph 6): I guess since 2005, Research Affliliates have been producing outstanding results.

Paragraph 7): I can't comment.

Paragraph 8): "Controlling risk than simply maximising their returns." Interest in Smart Beta started around 2007-2008, why not 2000? why not 1987? why only now?

Paragraph 9): (Leave for others to comment)

Paragraph 10): (Leave for others to comment) I don't think many Super Funds are using Smart Beta, please let me know if any do

That's my simple understanding (or lack of). DeepState, I awaited your contribution and views.

 

[The Falcon]

Yep, its an elegant concept, and no doubt, it offers on paper outperformance of dumb market weighted indices. But as an investor, we aren't looking for on paper performance, what matters to us is after tax performance. I offer as an example, Russell Australian Value ETF, ASX : RVL. At first blush, this ETF looks like it might offer outperformance for the lazy "value" type investor. It follows Russell's Australia High Value index, which is a rules based index comprised of companies that have low PE and strong medium term earnings growth.

Management fee is 35bps, which isn't bad and performance, net of fees (ie. after fees) has provided healthy outperformance over the last 3 years since inception. RVL accumulation 16.32% pa vs ASX300 13.30%.

ok, sounds pretty valuey to me, and I'm lazy but like to consider myself smarter than the herd...this ticks all the boxes! 3% pa outperformance for doing nothing? sign me up.

So the lazy value indexer is feeling pretty good and then tax time rolls around, he's done fairly well for himself and is in the top tax bracket. He gets his end of year tax statement from Russell, and sees that they have churned this bloody thing, RVL is distributing 6-9% and the lazy value bloke now has to deal with non discounted capital gains of 30-70%. In 12/13 almost 76% of the distribution was non discounted capital gains, and only 13% franked divis....compare this to his dumb mate who has just invested in VAS (ASX 300) index fund at 15bps, which distributions are 60-75% franked and there is bugger all, non discounted capital gain. I know who is happier come tax time.

RY, this is what I was referring to as not so smart beta, after initially being attracted to the headline numbers, a bit of digging around shows that all that glitters, in the hands of the investor post tax, is not gold. Rules based ETFs create lots of friction, which usually just damages after tax outcomes for investors...I read somewhere that Vanguard had to cut the rebalancing on their high yield fund from quarterly to twice annually, as the rebalancing was creating far too much non discounted capital gains.

Obviously its a much bigger subject than just this, but this is just one aspect that people need to consider and is often, conveniently from a managers perspective overlooked.

For mine, I have decided to stay with the dumb and cheap market weight ETFs (VAS and VTS) for my index component, and look for outperformance myself.....at least then I can control the tax events! I have some thoughts on the Small Cap tilt thing as well but have to get on a flight shortly so have to run...if I have time i'll post on that from Boracay, otherwise next week

 

[DeepState]

Hi DeepState
Can I offer any contribution? (I am really struggling to understand it - I am struggling to evaluate its merits or lack of merits from your view?)

Paragraph 1): Last sentence stating volatile markets and uncertain economic climate. This cycle has always happened. Smart Beta should have been "invented" or "used" or at least sold to the naive investor for many years. This is the first time I have heard of Smart Beta - I must be very naive.

Paragraph 3): Please help me. It shows my lack of understanding of how Fund Managers work. Especially when they need to re-balance their portfolios according to the Top ASX 50, ASX 100, ASX 300????

Paragraph 4): Alternative weighing schemes like volatility or dividends. I cannot understand how this can produce a better result. Some companies borrow to pay dividends (to artifically hold up a share price that doesn't deserve to be at the level?) I need some explanation please.

Paragraph 5): "to take advantage of perceived systematic biases or inefficencies market" OK Everyone thinks they can do this. Everyone except me since I am a beginner. The manager is "passive", so therefore it is cheaper than an "active" manager.

Paragraph 6): I guess since 2005, Research Affliliates have been producing outstanding results.

Paragraph 7): I can't comment.

Paragraph 8): "Controlling risk than simply maximising their returns." Interest in Smart Beta started around 2007-2008, why not 2000? why not 1987? why only now?

Paragraph 9): (Leave for others to comment)

Paragraph 10): (Leave for others to comment) I don't think many Super Funds are using Smart Beta, please let me know if any do

That's my simple understanding (or lack of). DeepState, I awaited your contribution and views.

Smart Beta is generally described as a rules determined way of building a portfolio to extract what are thought to be sources of persistent returns for reasons of risk bearing, illiquidity, behavioural bias, market dynamics... It is not seeking to profit from detailed knowledge of individual circumstances.

1. The concept of 'Smart Beta' is a repackaging of ideas that are 40+ years old. They are now offered as a middle road between index and fully active approaches which are cheap. Although rapidly growing, they are still a drop in the capital markets ocean. It has really taken off post GFC. It comes via the insto channel first as they are better resourced to grasp the merits of the idea and make a determination about their suitability.

3. The standard market indices like the S&P/ASX 200 are capitalization weighted. As a result, except for index reconstitutions (names coming in and out, rights issues...) portfolios indexed to them never need to be rebalanced. The argument follows that when a stock goes up in price relative to another, there is a fair chance that it has just become more expensive relative to another. If so, you should rebalance back towards where they were to some degree to capture this effect. The Smart Beta crowd argues that, given the indices do not allow for this, they are inefficient.

4. This is really tough. If the thread develops on alternative weighting schemes, we can get into it. However, from a wider sweep, if you rebalance to anything other than price movement (which does not need to be rebalanced), you can expect to profit versus the index before fees, t-cost and expenses are considered. That is, you will very likely beat an index alternative over time if you rebalance to something else. This something else can be just about anything: equal weights, weights that take into account relative volatilities, based on dividends / sales / cashflow / book value...or much more sophisticated. Rebalancing helps you to sell stocks whose price has moved up and buy those which have moved down. Because the market is characterized by excess price movement, this is expected to produce a superior outcome through time relative to the index alternative.

5. There is a lot of debate around this exact issue. This is called Smart Beta for a reason. Alpha is seeking to produce returns over and above those which can be generated from simple highly diversified portfolios which could be built without insight to specific situations. Smart Beta does not aim to make a return from knowledge of specific situations at all. Beta is the exposure to some bulk risk (say, equity) that is expected to yield a risk premium for bearing it. You buy this in pure form via index funds. Smart Beta is little twists on that theme that seeks to find certain premia that are more refined which produce an even better risk-reward tradeoff than bulk beta. Hence, it sits somewhere between passive and full-bore active. I'd say that it is an active fund seeking to beat the benchmarks via tilts to factors (eg dividends) that produce superior risk/reward outcomes to an index alternative.

6. Rob Arnott and Jason Hsu have been pushing the barrow for a while and have built an enormous business. Their approach is very simple and KTP could build the original in a few days (if that). You can find out more stuff from their website. They are the leading proponents of the practice in terms of AUM. However, there is fierce rivalry and debate of the type you might find here in ASF that goes on between participants. RAFI has good stuff on what is Smart Beta. If you want to get up the curve a bit more, then EdHec is a deeper source of info.

10. Smart beta is becoming more common now amongst insto. DFA is smart beta. RAFI offshoots also do this. Within many active managers are permanent style tilts which are actually Smart Beta. That is, they are charging you fees as if they are fully alpha, yet a large part of the outperformance is available cheaply via Smart Beta. In many cases, if you strip out the Smart Beta, you end up with nothing left over or value destruction. This is one reason why Smart Beta is gaining traction at the expense of active management. For example, State Super Financial Services uses it.

 

[systematic]

Any views?

Mine are:

As a concept
Nothing new here. It's factor tilting. Agree with your later comment about repackaging old ideas.

As a product
Possibly a bit too expensive for what you are getting. Due to number of holding etc some of these get close to closet indexing.

As an investment
Possibly a good one: for the person who (a) wants to go a little bit beyond "buying the market" in their equity investing, but (b) do not want to touch active investing themselves.

 

[DeepState]

RY, this is what I was referring to as not so smart beta, after initially being attracted to the headline numbers, a bit of digging around shows that all that glitters, in the hands of the investor post tax, is not gold. Rules based ETFs create lots of friction, which usually just damages after tax outcomes for investors...I read somewhere that Vanguard had to cut the rebalancing on their high yield fund from quarterly to twice annually, as the rebalancing was creating far too much non discounted capital gains.

Obviously its a much bigger subject than just this, but this is just one aspect that people need to consider and is often, conveniently from a managers perspective overlooked.

For mine, I have decided to stay with the dumb and cheap market weight ETFs (VAS and VTS) for my index component, and look for outperformance myself.....at least then I can control the tax events! I have some thoughts on the Small Cap tilt thing as well but have to get on a flight shortly so have to run...if I have time i'll post on that from Boracay, otherwise next week

I had to look up Boracay. Nice!

On the same page as you. The net of tax outcome is what you want. If you are investing on max tax rate, it's hard to justify any turnover at all in capital account assets, for the most part. That Russell ETF probably torched a stack of franking as well. You can control tax to a large degree (except the law). Everything else is just a promise.

Have a great time.

 

[Faramir]

Thank you DeepState. Thank you for your efforts. It will take me a while to take all of the info in but you have helped immensely. I just realised that this thread is in response to a Myer Holdings (MYR) thread. I am not surprised that some 'active' fund managers claim that they are actively trying to beat the index and charging for it but they are 'cheating' by using Smart Beta. If I ever buy an Index Fund, I have to be careful. Not just look at returns results. I must look at Fees, methods that they use. I wonder if they would disclosed the fact that they use Smart Beta? So much information for me to take in.

 

[luutzu]

Woah!

Quite a racket the finance and investment industry got going.

It's incredible how people openly admit they aren't that good at their job, are only figuring out ways to not stand out among the herd, then charge cart loads for "expertise" in getting around the same rate as the entire market would.

Can't wait for Smart Beta Max

 

[TPI]

I like the smart beta concept. I've been reading the blog posts/articles from the Research Affiliates website in the last 6 months or so and it all seems quite logical to me. Though I haven't read any opposing views on this approach yet. My own share portfolios are DIY versions of a smart beta portfolio, with a few different rules and fundamental tilts created by myself based on my reading of what works in investing.

 

[luutzu]

I like the smart beta concept. I've been reading the blog posts/articles from the Research Affiliates website in the last 6 months or so and it all seems quite logical to me. Though I haven't read any opposing views on this approach yet. My own share portfolios are DIY versions of a smart beta portfolio, with a few different rules and fundamental tilts created by myself based on my reading of what works in investing.

Whatever happen to aligning one's portfolio with stocks that actually earn money as measured by such things as it actually earning money from doing real businesses?

I really feel sorry for people trying to sell something to institutions that require them to do real work.

Anyway... let's not complain too much when others are trying to help you

 

[DeepState]

I like the smart beta concept. I've been reading the blog posts/articles from the Research Affiliates website in the last 6 months or so and it all seems quite logical to me. Though I haven't read any opposing views on this approach yet. My own share portfolios are DIY versions of a smart beta portfolio, with a few different rules and fundamental tilts created by myself based on my reading of what works in investing.

Good stuff. I like their (RAFI) work as well. Please note...you need to rebalance.

Main detractions:
1. What seems smart beta might actually be stupid beta.
2. It is an active tilt from index despite trying to ignore the index. As a result, it cannot absorb an enormous amount of money before the idea asymptotes to zero.

What's new?

 

[TPI]

I think for most part-time/non-professional stockmarket investors who are fundamentally inclined, there is no choice but to use some form of ‘’smart beta’’ or ‘’factor tilting’’, or whatever label you want to give it, to move the odds of positive performance in your favour.

If you have limited time and limited depth of knowledge/experience/analytical skills then to me it makes sense to at least tilt your portfolio towards things that have been shown to give some sort of performance edge over the long-term.

I think most of us do this to some extent even if unintentionally, but perhaps sometimes tilting towards the wrong areas, including myself.

Personally, with my investing approach I rarely read annual reports in great detail, usually only the first few pages, and rarely break down financial statements in great detail either, due largely to the limitations mentioned above.

With this in mind, if I am going to pick stocks by myself, I need to at least have a few portfolio construction and stock picking rules that are going to help me tilt the odds in my favour, along with some basic level of diversification to protect me from bad stock picks.

In broad terms for me this means things like:

Low volatility stocks (which usually leads me to an overall large and mid-cap tilt)

Dividend yield and dividend growth

Low P/E and/or discount to some measure of intrinsic value

High ROE

Low debt/equity and/or high interest cover

High margins

Positive earnings and earnings growth

Use of some leverage in the portfolio

About 10-15 stocks in the portfolio (a balance between concentration and diversification)

Notwithstanding the above, I still have some small cap and speculative stocks as well, though Research Affiliates recently posted an article saying that the small cap premium no longer exists, which I found interesting…

And tax considerations are always important, so outside super I tend to have a much lower turnover portfolio than inside super, and my small cap and speculative tilts are greater within super and with no leverage used here.

And I usually try and top-up on under-represented stocks in dips rather than sell-down others to rebalance, though I did sell down some bank shares recently as I felt I was too top heavy here.

In terms of what is and isn't smart beta, from what I read low volatility, value and market factors are considered smart beta (which my criteria includes). Momentum is another but I don't use this very much in my process just yet. And various quality definitions, presumably like high ROE, low D/E, high interest cover, high margins etc., are not - nonetheless it still makes sense to me so I still use these to filter through stocks!

And there is also a illiquidity premium which is meant to be significant I think, but I'm not quite sure what this refers to?

 

[TPI]

Whatever happen to aligning one's portfolio with stocks that actually earn money as measured by such things as it actually earning money from doing real businesses?

Nothing, that's your smart beta, just how much time and depth of study you devote to establishing what these businesses are will determine whether it is now a more active process.

And whether the academic papers say it is smart beta or dumb beta is another thing!

 

[DeepState]

And there is also a illiquidity premium which is meant to be significant I think, but I'm not quite sure what this refers to?

If an asset is not tradable readily, it will cost you more to get in and out. Your opportunity costs are also high because it prevents you from readily deploying the capital to higher return purposes due to the transaction frictions. Hence, you need to charge a higher return for transaction costs and for opportunity costs even in a regular steady state situation. If these assets are purchased with leverage or are otherwise affected by cyclic demand, then there is even more risk because you can't get out when you are most likely to need to get out. A greater margin of safety is required again.

For these reasons, two assets, one listed and the other not will trade at different prices despite identical underlying operations. You can infer the price from looking at various similar securities whose underlying is similar but whose market characteristics are different for market micro-structure reasons. This occurs even in US treasuries. The difference is the (il)liquidity premium.

 

[KnowThePast]

My limited experience tells me that the concept makes a lot of sense, and seems like a relatively easy thing to do. After all, you only need to make a couple of decisions to "tilt" your portfolio in areas where you are confidence of better performance.

In practice, it rarely happens that way. Meddling with an index, or automated strategy, tends to worsen the performance, not improve it.

I remember reading a few books on automated strategies that talked about this phenomenon as well, but I don't remember which. The reasons given, I think were psychological, and exclusion of "unlikely outliers".

My 2c.

 

[luutzu]

"The way that can be defined is not the Way."
- Lao Tzu

 

[TPI]

If an asset is not tradable readily, it will cost you more to get in and out. Your opportunity costs are also high because it prevents you from readily deploying the capital to higher return purposes due to the transaction frictions. Hence, you need to charge a higher return for transaction costs and for opportunity costs even in a regular steady state situation. If these assets are purchased with leverage or are otherwise affected by cyclic demand, then there is even more risk because you can't get out when you are most likely to need to get out. A greater margin of safety is required again.

For these reasons, two assets, one listed and the other not will trade at different prices despite identical underlying operations. You can infer the price from looking at various similar securities whose underlying is similar but whose market characteristics are different for market micro-structure reasons. This occurs even in US treasuries. The difference is the (il)liquidity premium.

Thanks RY, for stocks does this apply to stocks that are liquid vs ones that are less so eg. small caps with low volumes and large bid/offer spreads? So is it better to invest in more liquid stocks to capture this premium or the opposite?

 

[DeepState]

Thanks RY, for stocks does this apply to stocks that are liquid vs ones that are less so eg. small caps with low volumes and large bid/offer spreads? So is it better to invest in more liquid stocks to capture this premium or the opposite?

Here's some US data:



The liquidity premium exists across all size bands. That means that you do not have to tilt to small caps to get the effect. You can, but you do not have to.

Interesting to see the small cap premium in the data. Interesting also that the relationship does not exist for the most liquid quartile and is monotonically in the opposing direction.

Much of this smart beta stuff is actually finding rewarded risk premia. It's not that the returns are there for nothing. It's just that you may be in a better position to absorb these risks than others in the market and hence get the rewards that come with it. Ultimately, the secondary market is a risk transfer mechanism.

Seriously impressed by the thought you have put into this BTW.

 

[tech/a]

What sort of return is expected from this?
Which is how much more than a return claimed
Without it?

How would this affect someone who runs their own SMSF.

Why would I care?

 

[DeepState]

What sort of return is expected from this?
Which is how much more than a return claimed
Without it?

How would this affect someone who runs their own SMSF.

Why would I care?

The difference between the first and fourth columns highlights the return you can get from buying illiquid stocks and selling the liquid ones, sorted by size/capitalization quadrant. The table reports that this one idea has generated somewhere between 3-12% per annum pre t-cost, borrow and finance margin, depending on what you would have wanted to buy and sell assuming you held a fully self financing portfolio which is long $X of illiquids and short $X of liquids against it. That is without (net) leverage. The long-short portfolio would have lower risk than one just invested in equities without leverage. Long short portfolio returns would be added to the cash rate to obtain a sense of total return from a dollar-matched, hedged, portfolio.

There are many ways to utilize this idea, if you wanted to do so. A long-short portfolio rebalanced annually is only one. It can be applied to an SMSF or any other type of investment vehicle. Either outright or as part of a wider investment strategy.

I don't know why you would care or not. It's up to you. It's only money.