Sign in     Like us on Facebook Follow us on Twitter Watch us on YouTube

News and Blog

Join 5000 other sharemarket traders for regular blog updates!

Browse to a category

Blog Search

ETF’s that ate the world

Interesting graphic from the Investment Company Institute showing the flows of cash into index ETF’s and out of actively managed funds.


FTSE 100 All Time High

It is interesting to note that the FTSE 100 has made a new all time high (yet another market doing better than our own). The FTSE presents an interesting lesson in how markets can behave and how long it can take for a market to recover. The FTSE made a new all time highest close in December 1999 but it didn’t breach that mark for another 15 years when it briefly flirted with new highs for a few weeks before falling again.  It then took 9 months to recover to the new highest close. What is interesting is the amount of time it took to move beyond the 1999 mark. This has undoubted implications for index traders, particularly those who buy ETF’s. Index ETFs have a few strong selling points – you will generally always beat active managers over the long term and you will do so at very low cost.  However, it can be a long time between drinks if the index decides to go nowhere for decades.

It will be interesting to see if participants hold their nerve.

100 dd

More ETF Magic

About the only joyous thing about Christmas is that there is a marked slowdown in the amount of junk  mail that hits my inbox. And to be honest I do miss the emails telling me that people have the ideal job for or that Svetlana from Kokshetau in Kazakhstan is desperate to send me pictures of her recently shorn yak. Paradoxically I have not seen a slowdown in the number of emails extolling the virtues; nay magic of index ETF’s. This got me thinking about the nature of indices and their construction. In simplest terms an index is simply an aggregation of stocks design to represent a sector of the market. However, stocks are not simply lumped into a group and passed off as an index, an index can be either price weighted such as the Dow or market value weighted index such as the S&P500.  With the Dow the weighting of each component is a function of its price whereas with the S&P500 components are weighted according to total value of their outstanding shares. Each method of construction results in a different end product.

Because we have two mechanisms of solving the same problem we can compare and contrast what each solution looks like when plotted against one another. This might seem to be a somewhat academic problem and in many ways it is is except when you consider that when being told to buy an international ETF investors are directed towards SPY which is a representation of the S&P500 –  a market weighted index. You might think so what but remember this is not the only way to measure an index and that the folks who create ETFs are nothing if not inventive. As such there is a price weighted version of the S&P500  known as the Guggenheim S&P 500® Equal Weight ETF or RSP

It is therefore easy to compare the performance of these two instruments –


For the period being considered it appears as if the equal weighted version has given the market weighted ETF a bit of a belting. This raises the question as to why do dislocations occur between the performance of two indices constructed using different methodologies and the blindingly obvious answer is that it is the different methodologies account for the different performance. Granted they hold the same stocks but they hold them in different ratios. For example SPY’s largest holding is Apple which makes up 3.12% of its portfolio whereas its holding in RSP is capped at 0.21%. This means that when Apple does well SPY will do well but when it does poorly it will drag the performance down. Equal weighted indices are to some extent insulated against the travails of a single stock.

The existence of instruments such as RSP introduce a layer of complexity for ETF traders since they offer a variation on the theme of index investing that may or may not suit them. Unfortunately, for those who pump magic index ETF systems the world is a little more nuanced than they think it is.


Another Day Another ETF System

It does seem as if every time I open my junk mail folder I get another invocation to partake of yet another magic trading system. I should note that this sort of thing has been around forever, the only thing that changes is the vehicle  which alters in order to capitalise on whatever topic is hot. Or whatever is perceived to be magical. With the need for a magic system foremost in my mind I wondered how long it would take me to build a magic system that could trade the major Local ETF – STW.

Lets just say the kettle hadn’t had time to boil by the time I was finished. The results of the system are shown below.


The rules are simple – entry is a new 52 week high and the exit is a new 52 week low. The figures above represent buying and selling on the opening of the week immediately following the generation of the signal. You will note that I have been profoundly lazy since I have not included any money management in the analysis. These are effectively block trades – you buy a chunk on the signal and then wait for a signal in the opposite direction. If I get some free time I might be a little bit more sophisticated in my analysis but I just wanted a proof of concept to see what sort of results would be generated by being very blunt and letting the market do the work. But you will notice there is nothing magical nor particular sophisticated about this. As I have said before any robust trend following method will work with any instrument.



There Is Nothing New Under The Sun

One of the joys about having been around since Adam was a pup is that you get to see lots of things come and go. On a regular basis you get to see what the new flavour of the month is and you quickly learn that there are no good or bad instruments just good and bad trading ideas. Over the past little while I have been watching the excitement around ETF’s reach some sort of gushing crescendo as apparently they are the saviour of every long term investor in the world. According to the hype all you have to do is buy and ETF and your retirement is assured. Unfortunately, one of the things you also learn with experience is that the world is a little more nuanced than that. Much of what is written about ETFs is correct – they are a simple low cost way to gain exposure to the market and their growth has been extraordinary as can be seen in the chart below.

ETF Global Assets

Source: Financial Times

When we come to trade an instrument there are many facets to our decision making. Naturally we need some sort of trigger to spur us into action – this is generally some form of trading signal that has been derived from the price of the instrument itself. However, part of the trading process also involves deciding whether the environment for trading will be friendly towards us and enable us to enact all aspects of our plan in a friction less manner. It is at this point that we run into the very diverse ecology that is the ETF landscape.

Not all ETFs are created equal

In the trading process getting into a trade is generally fairly easy and the ease of entry reflects the relative lack of importance of entry as a requirement to successful trading. It is the ability to exit a trade with ease that over time determines our survivability as a trader. In setting benchmarks for the ease with which we can move into or out of positions we generally set a liquidity high water mark that has to be passed before we consider a vehicle viable for investment. This can be any number and in looking at ETFs in more depth I came across a report by a group known as Stockspot who produced a very handy little guide to local ETFs and I urge everyone to have a look at it. In their guide they rank ETFs on a variety of criteria one of which is liquidity and they set a benchmark of $500,000 per day as the minimum threshold. I find it hard to argue with this hurdle as a barrier to entry. However, I want to take a step back and look at liquidity more from the perspective of a business risk and for this I looked at AUM. From experience and observation it has hard to run a fund below a certain AUM figure – the expenses involved just don’t make it worthwhile and ETFs do face a business risk from the perspective of the mangers just quitting, which can and does happen. In addition to this you face an accessory risk with these exchange traded products and it comes from the manager inability to adequately manage their market making responsibilities.

In the tables below I have ranked local ETFs on the basis of funds under management. If we use the general benchmark of $250M as a minimum then you can see that very few local ETFs hit this benchmark. The majority are way below it and a reasonable number would seem from my outside perspective to not be viable. (click for full sized image)






Beats an Active Manager

This point is true but with certain limitations – index based ETFs will beat on average a population of active managers. It is often stated that almost 86% of all active managers in the world under perform their benchmark index. Therefore, it seems to be logical to assume that investing in a low cost manner in the index itself would be more efficient and with regard to ETFs this point is true.

However, there is a small point that most miss. The ETF itself also under performs the index – this little point is glossed over by everyone. This occurs because of the nature of fees as they impact upon the growth of the ETF. Consider the chart below which compares the S&P/ASX 200 Total Return Index with the popular STW – there is a gap in their performance. This gap is a reflection of the cost of having someone else assemble a portfolio for you that matches the index.

STW vs ASX 200 TR Google

This dislocation in performance persists throughout all market phases. This dislocation is known as tracking error. This error is never zero and it may be positive or negative.

What you see is not always what you get

The notion of a simply buying and forgetting is not the path to riches that is portrayed. Imagine the following scenario, I take a view that is long the Nikkei and I want to exploit this via an ETF so I scan the available local vehicles and I come across IJP which I buy. Sometime later I look at the performance of the Nikkei and think I must be doing brilliantly and therefore my ETF must be doing brilliantly as well. Unfortunately, it’s not that easy.


If I had held my ETF from 2013 to 2015 I would have underperformed the market by a reasonable margin. The tracking error in this instance has come from the impact of currency fluctuations. During this period the AUD was broadly stronger against the JPY and this impacted the return on the fund. Over recent months this relationship has reversed and the fund has fared better than the index.

You don’t always just have one trade.

I am fond of saying that when you undertake any trade you are actually undertaking two trades. The first is the underlying instrument – the second is a trade on you. The metric for judging this trade is how well you followed your plan. However, with ETFs that track foreign indices the initial trade is a little bit more complex because of the impact of currencies.

For example consider the comparison between IVV.US and IVV.AU. These are essentially the same instrument but one is quoted in the US and one locally.



Despite the similarity you can see substantially different performance over the long term.

However, this brings us to a further problem which is more of a general trading problem and it has to do with the way results are presented. In presenting results the starting date is everything. In the chart above I have presented as long as view as possible. But if I compress the view to five years something interesting happens as can be seen below.


The relative performance changes with the domestic entity having superior performance and this new start date reflects the collapse in the AUD from 2013. The primary trade is simply being long the index; the secondary trade is the currency fluctuations that this trade naturally comes with.

You need to be careful in understanding where the performance is coming from. As a general rule you can tell when performance figures have been fudged because the system or idea seems to make money from day one. In reality what happens with trading systems is that they should all drawdown first and then start to climb. This occurs because well designed systems take their initial losses quickly; it then takes some time for the gains to accumulate.

It has oil in the name therefore…..

If I look at a list of the most popular ETFs by traded volume within the top five you will generally find  United States Oil Fund (USO). This is by far the most popular vehicle by which people seek to gain exposure to the oil price and I am going to make one of the usual dogmatic generalised statements I am prone to make. Most of the people trading this vehicle have made an error and the error occurs in not understanding what they have bought and what it actually tracks. If I look at the definition of USO I see the following –

The United States Oil Fund holds near-month NYMEX futures contracts on WTI crude oil.


USO, among the largest and most liquid oil“a great vehicle for riding short-term moves in crude prices” ETPs available, delivers its exposure to oil using near-month futures. USO’s huge asset base waves away closure risk, and its massive liquidity makes trading a snap. USO gets exposure to oil using derivatives, like all oil ETPs. Derivative returns can vary greatly from spot oil prices, but spot oil is uninvestable.

USO holds front-month futures contracts on WTI, rolling into the next contract every month, just like our segment benchmark. This method is particularly sensitive to short-term changes in spot prices, but can also result in heavy roll costs. That makes USO a great vehicle for riding short-term moves in crude prices, but long-term holders may want to look at other options.

There are two parts to this definition that need to be explored – the first is simple hyperbole and it is the bit that the majority of traders will focus on. In particular the following is attractive to traders –

USO’s huge asset base waves away closure risk, and its massive liquidity makes trading a snap. USO gets exposure to oil using derivatives, like all oil ETPs. Derivative returns can vary greatly from spot oil prices, but spot oil is uninvestable.

To get a sense of the problem with USO we first need to see if there is a problem, after all the marketing tells me that it is a great vehicle for riding short term moves in oil. Consider the chart below which compares the yearly return figures in the spot price of crude with those of USO.



As you can see there is a substantial difference and this difference is related to the second part of USO’s definition –

USO holds front-month futures contracts on WTI, rolling into the next contract every month, just like our segment benchmark. This method is particularly sensitive to short-term changes in spot prices, but can also result in heavy roll costs. That makes USO a great vehicle for riding short-term moves in crude prices, but long-term holders may want to look at other options.

USO holds front month contracts when these expire they have to be rolled over and in a market state known as contango these trades end up chasing price. To understand the idea of contango consider the prices below:

Futures Price

These are a series of prices for crude, as you can see the further out in time we go the higher the prices become. Contango is a natural feature of markets and doesn’t really imply anything other than market perceptions. Traders often think that contango is the normal state of affairs or reflects a normal market – this is not true. The issue for USO is a structural one, when a contract expires the next contract needs to be bought and this process is known as rolling, this rolling comes at a cost.

This lack of understanding of the mechanics of futures trading results in traders buying the wrong ETF for the right reason. In reality traders looking for longer term exposure should have opted for USL which looks to try and deal with the problem by only rolling one-twelfth of its portfolio. Alternatively, they could have opted for something a little more sophisticated such as trading XOP which handily outperforms both major oil ETFs as can be seen below. However, such a decision is dependent upon understanding the structural limitations of what you are trading.

Why has it not gone up three times as much?

The final point I want to make about ETFs is in regard to the use of leveraged ETFs. Traditionally these are ETFs whose name is prefixed by 2X, 3X or Ultra. The perception is that these vehicles will deliver a multiple of the return being generated by the underlying instrument and they can be either bullish or bearish in nature. The problem for traders comes in with the assumption that these tools can be held for more than a single day. This misunderstanding comes from failing to come to grips with what is actually meant by being leveraged and how the return is calculated. The returns are calculated on a daily basis therefore the assets of the fund are constantly being rebalanced so unless an instruments trends in the same direction day after day after day the gains made are eroded.

We arc back to a familiar refrain – if you don’t know what it is don’t trade it.

The wash up

The basic question that any trader needs to ask with regard to using any instrument is does it suit my view/approach/ philosophy and if it does can it be successfully integrated into my arsenal of instruments/techniques. As I stated in the outset there are no good or bad instruments only good or bad decisions and with instruments such as ETFs these decisions are often made on the hype of the instrument as opposed to its actual utility. For example much of the marketing around ETFs seems to fall into one of two categories. They are magic and will solve all your problems you just buy them and forget them and because they are magic you need a magic system to trade them.

With regard to the first point ETFs do fill a void in the market, they offer a low cost way of tracking an instrument when it is trending. There are limitations to this and there are potential sources of problems but if you are aware of them then they shouldn’t present any great issue. This leads to the second point about needing special systems to trade ETFs; all you need is to be able to follow a trend and this can be done with everything from a moving average to a series of new 52 week highs. Nothing special, nothing complicated.








ASX 200 Shares Distribution Of Returns

Recently I had a look at the distribution of returns for a system I run – the aim of this was to give a sense of how systems trading works. The rules of systems trading are very simple and are based around the concept of ride the losses, pump the winners and allow time to do its thing. Following on from this I thought it might be interesting to have a look at how the individual returns for shares that make up the ASX 200 were arranged and to see if this had any lessons hidden in the data. From my perspective it did prove to be interesting but then again I find the structural data of the market interesting.  The first thing I needed to do was to calculate the 10 year annualised return of each stock. There are a few caveats with this sort of exercise and they need to be understood before interpreting the data.

1.This is the S&P/ASX 200 as it stands not the S&P/ASX 200 since inception. As a result survivor bias is an issue.

2. Annualised returns can be misleading. It is possible for a stock to do nothing and then have an explosive single year. This single year distorts the data upwards.

3.Likewise it is possible for a stock have an appalling year and have the average result dragged down.

4.Some stocks in the S&PASX 200 have very limited data with some trading for less than 2 years – this gives a distorted appearance to their returns.

5.Do not assume that this is the return generated by the instrument every year. Pay attention to points 2 and 3.

When I tabulated the data I got the following –


Upon initially viewing the data in this raw format the thing that struck me was the amount of rubbish that is listed locally and which for some reason qualifies to be in the markets benchmark index. Several of these companies have limited price history and some have applying price discovery as evidenced by how illiquid they appear. What is also obvious is that many listings have a long term negative pay-off. However, as stated we need to be a little bit careful when making this statement because of the way this data has been calculated. What did surprise me was that I thought that more stocks would have a long term negative return and then I remembered my own warning about survivor bias. The S&P/ASX 200 is turned over regularly so the dogs are flushed from the system and new stocks are added in an attempt to pump the index. My guess is that only a handful of stocks from the original listing of the S&P/ASX 200 still exist (note to self – research project for later) This is why indices have an upward bias.

This data by itself is interesting but doesn’t really convey much information so I decided to have a look at the frequency of returns for the index and this yielded the chart below.


Interestingly, there are more positive returns than negative returns – my explanation for this is once again its a problem associated with survivor bias. However, it is always interesting to see the clustering of results around a given point. In this set of data the mean return is 13.65% with the median being 10.25%. I then started to segment the data into various bins to see how different segments looked. The S&P/ASX 20 showed me an interesting pattern. I did  this because I made the assumption that the S&P/ASX 20 would be the most relatively stable collection of stocks and it gave the following distribution of returns.

top 20

What is interesting to me is how average the returns for the major banks are over this period. Granted the starting point of the data does include pre-GFC highs but this is also a problem for all stocks in this sample. It is also important because with the failing of BHP and RIO advisor’s tend to overweight portfolios with these stocks. Breaking the data into segments lead me to splitting the data into the top 100 of the index and the bottom 100 in terms of market capitalisation. Doing so confirmed something I already knew. Larger cap issues have less room for outsized returns. The top 100 generated an average return of 11.81% whereas the second tranche generated a return of 15.56%.

This is a significant difference and leads me postulate if given the various assumptions inherent in the data if it is not better for investors/traders to hunt outsized the major index components. I know this to be true from my own systems testing and stock selection criteria. The reasons for this are many but my assumption is that it is a function of leverage and stagnation. Smaller issues have a lower market cap and make up the bottom tier of stocks. This leverage enables larger gains. The stagnation element is related to something I wrote last week. Companies have periods of explosive growth and then they begin to stagnate. Larger companies that have extended history are largely moribund in their thinking and business strategy. For example, if you ran one of the big four banks you would know that you would make extraordinary profits by being ordinary. Given that this is your default setting why would you do anything to change this. It is not in your interest as a senior executive or company director to do anything out of the ordinary. If by some whim of fate you were suddenly made CEO of one of the major banks your instruction ot your staff every Monday morning would be to do exactly the same things you did lat week, the week before that and the week before that. You could then collect your $18,000,000 plus salary and bugger off and play golf for the remaining four and half days of the week safe in the knowledge that the money would simply roll in.

From the perspective of the average trader this data does convey some lessons, most prominent of these is that size matters in limiting returns. This data merely confirms what systems traders have known for decades. However, I think there is something more important in being granular in your view of data such as this. It actually tells a lot about the structure of the market you are trading in. Too often both investors and traders merely look an index and assume that its components are probably homogeneous and therefore not worthy of further examination. In fact I have come across many in the advisory index who were unaware that stocks were dropped from and index and then replaced hence the upward drift of an index. They just merely assumed that the index always went up because of some magical reason. What is worse is they also assumed that every stock in that index went up as well. Looking deeply at data removes these preconceptions and puts you int touch with what the nature of the market is. It also and most importantly makes certain that your trading system is performing optimally by giving you an edge in the stocks you look at.

Australian History 101

When I was a teenager I shifted schools. I moved from an inner city school that taught a diverse range of history’s such as Japanese, Chinese and a range of European focused history’s to a smaller regional school that for some reason taught Australian history. Which I found as dull as dishwater, after all not much can happen in 200 years. I can give you my potted version of Australian history in a few points.

1. The Poms and the Frogs turn up. The Frogs have a quick look around inner Sydney and say in French of course – lets bugger off back to Tahiti where it is warm and there are palm tree’s.

2. The Poms stay and lob a metric shit tonne of convicts on the place because they had run out of space at home for people who had committed heinous crimes like trying to feed themselves. Intriguingly, a bunch of in bred ponces in perfumed wigs had not worked out that if you treated the bulk of your population as filth then you might run into one or two social problems.

3. We begin 200 hundred years of screwing up relationships with the people who were here first. Again, see reference to perfumed ponces for the quality of decisions you could have expected on this matter. A quality of decision that seems to persist to this day.

4. We work out that sheep breed like Catholic rabbits in the local climate and set in train the industrial policy of shipping things to foreigners and buying them back  in a slightly modified form at exorbitant prices.

5. Rum Rebellion – a bunch of soldiers get pissed and apparently this is a defining moment. If so then there have been an infinite number of defining moments in our history.

6. We discover gold and briefly towns such as Bendigo are amongst the richest in the world. This doesn’t last long because all the wealth disappears, again setting in train an intergenerational policy of shifting wealth off shore. Good to see the tradition continues.

7. A bunch of miners get pissed and punch on with the local cops- apparently this is one again a defining moment.

8. Ned Kelly – an idiot wears a bucket on his head and fatally shoots three policemen. For some this is a matter of national pride to be celebrated along with tattooing the southern cross on your forehead.

9. Federation – why have separate governments when you can add another layer on top and then add local councils as well. Happy days for everyone.

10. Gallipoli – we turn up on someone else’s door step and they are quite rightly annoyed and belt the shit out of us. The entire thing is cock up from start to finish and confirms in the national psyche the need to celebrate failure. We celebrate the event by shipping thousands of pissed, poorly behaved Australians back to Gallipoli once a year thereby pissing the Turks off once again.

Dont think I missed anything.

However, in amongst all of this detritus I did learn something. It is better to sell shovels to a miner than to be a miner. With all the collective breathlessness currently doing the rounds regarding ETF’s it is necessary to take a deep breath and look at where and how you generate out performance in the market. There are a few points about ETF’s, particularly index ETF’s such as STW that are manifestly true and in no particular order they are.

1. They allow a broad market view that can enable investors to take advantage of long term moves in equities.

2. They offer a cheap solution for the management of compulsory savings schemes such as superannuation. In fact if you simply used index based tools the performance of all superfunds would probably lift dramatically.

3. They do allow simple means of exposure to other markets.

4. They are not magic. They can help average investors but they cannot save them from themselves.

From my perspective there are two key points. Firstly, if you are going to trade index ETF’s for outperformance then you need to select the correct one. Secondly, the natural question that arises is, are they the best vehicle for taking advantage of long term trends in equities markets?

With regard to the first issue most point to the use of STW as the vehicle of choice. STW is an index ETF that mimics the S&P/ASX 200 index and it would not be my first choice of vehicle. Over the past few years STW has underperformed ILC, which is an index ETF that covers the S&P/ASX 20. This underperformance can be seen below.

Screen Shot 2015-10-19 at 8.11.57 am

The reason for this differential is simple. When markets are recovering from sustained falls money moves first into what are perceived as quality or blue chip issues and this has been the situation locally. Fund managers have been loathe to buy outside of this criteria to any great degree. This relative performance can be seen in the performance of the S&P/ASX 200 versus the S&P/ASX 20. One is years off its all time high whereas the other has nudged at its previous record highs. Comparing to index ETF;s does not off course answer the question as to where to achieve outperformance in a rising market. To do that it is necessary to come back to my earlier statement about it is better to sell a miner a shovel than to be a miner. In the chart below I have compared the performance of STW to that of ASX – the listed entity that controls our local stock exchange.

Screen Shot 2015-10-19 at 8.12.44 am

As can be seen, being involved in the business that controls an exchange is more profitable than being involved in one of the metrics that measures the performance of entities within that exchange. So if I were looking for a long term instrument that moves strongly when markets are trending I would opt for the business with the most to gain both in reality and perception from that bull market.





General Advice Warning

The Trading Game Pty Ltd (ACN: 099 576 253) is an AFSL holder (Licence no: 468163). This information is correct at the time of publishing and may not be reproduced without formal permission. It is of a general nature and does not take into account your objectives, financial situation or needs. Before acting on any of the information you should consider its appropriateness, having regard to your own objectives, financial situation and needs.