Continue reading »]]>

According to (Elroy Dimson, 2010) the “conventional view Is that, over the long run, corporate earnings will constitute a roughly constant share of national income, and so dividends out to grow at a similar rate to the overall economy. This suggests that fast-growing economies will experience higher growth in real dividends, and hence higher stock returns”.

Unfortunately, empirical evidence showing positive relationships between economic growth and sharemarket returns has not been as strong as many have expected and there are many examples, where high economic growth has coincided with terrible sharemarket returns. Emerging Markets and China are possibly one of the better examples of this relationship breakdown in recent years.

The purpose of this article is to revisit the relationship between economic growth and sharemarket returns. Whilst much of the literature, including on this subject focuses on the US, the data presented in this paper also includes Australia and United Kingdom. Why Australia and United? … A couple of reasons and both somewhat naïve. Firstly, they are spread geographically with different major trading partners, and secondly, they were the regions which enabled easiest access to long term data.

For the analysis that follows, all equity market returns are accessed from MSCI database, are calculated in local currency (i.e. AUD for Australia, USD for USA, and Sterling for UK), and data series commence 31/12/1969. The specific indices are MSCI Australia GR, MSCI UK GR, and MSCI USA GR.

Real GDP figures come from each country’s central bank data sources. That is, Reserve Bank of Australia, Bank of England, and Federal Reserve. Please note, at the time of writing Real GDP figures were not available for the December 2016 quarter for either Australia or USA.

Let’s start with a direct comparison of equity returns and real GDP using quarterly data for each country. Figures 1,2, and 3 show the results for Australia, UK, and USA, and only USA shows any relationship. Both Australia and UK show no relationship between equity returns and the same quarter Real GDP results at all…this is supported by the flat trend lines and the R^{2} result being close to 0.

The USA Trend suggests that for every additional 1% in quarterly Real GDP Growth it correlates with an additional ~2.2% in quarterly equity market returns. With an R^{2} of less than 5%, Real GDP does produce a great deal of explanatory power of equity returns, but the slope s highly statistical significant given it’s t-statistic is 3.04.

This suggests that of the three countries, if you can predict the quarterly GDP in advance then it might provide a slight edge in the USA only … it doesn’t necessarily help in Australia or the UK.

Figure 4 – Regression results for Figure 3’s trend line

Source: Delta Research & Advisory

For many the potential spurious relationship between quarterly equity returns and Real GDP is unsurprising because the belief is that the sharemarket is more of a leading indicator, meaning that its direction is a forward predictor of future economic growth. The initial test for this is done by comparing the last period’s equity market returns to the current period real GDP.

The charts and results are shown in Figures 5,6, and 7. Once again, only the USA show a statistically significant trendline suggesting that last quarter’s equity market return might be a leading indicator for this quarter’s Real GDP. That said, the US Equity market return only explains around 10% of the variance in the next quarter’s Real GDP so there is a lot of unexplained variance due to other factors, which should not be surprising.

Neither Australia nor UK show a statistically significant relationship for last quarter’s sharemarket returns being predictive of this quarter’s Real GDP result.

Source: Delta Research & Advisory

So far we have looked at quarterly data only and many might argue that it is longer run economic growth that is important to sharemarket returns because quarter-to-quarter can be potentially meaningless due to volatility and the associated uncertainty or potential lack of obvious trend.

Figures 8,9, and 10 show the relationship between annual returns and annual Real GDP and the only significant trend is for the USA. The relationship between annual equity returns and annual Real GDP appears insignificant for both Australia and the UK.

Source: Delta Research & Advisory

Interestingly compared to the quarterly results from Figures 3 and 4, there is greater explanatory power in the regression model. Whilst quarterly Real GDP explained around 10% of the variability of quarterly equity returns in USA, annual Real GDP explained more than 28% of the annual equity market variability in USA between 1970 and 2015.

These results show only a stronger relationship over longer term, once again, for USA.

So is this year’s sharemarket return likely to provide an indication of next year’s Real GDP Growth? At the quarterly level, there appeared to be potential evidence suggesting the sharemarket might be a leading indicator in the USA only but at the annual level? It is the reverse result. That is, the relationship between lagged annual equity returns and real GDP growth is strong for Australia and the UK and weak for USA…refer Charts 12 through to 14.

In Australia, for every 10% in annual equity returns has meant next year’s Real GDP growth has been 0.383% higher. For the UK, that relationship is similar whereby every 10% in annual returns meant it’s Real GDP was 0.312% higher the following year. Still with R2 of only ~23% and ~15% respectively, there is still a lot of unexplained Real GDP variability as you would expect.

Source: Delta Research & Advisory

The final analysis looks at 5 Year data. Statistical tests on this small sample size can be fraught with danger so a table of Real GDP and Equity Returns is presented in Figure 15 below. As mentioned above, all equity returns are calculated in local currency terms; red cells indicate below average results and green cells indicate above average results.

Overall there does appear to be some potential patterns. The Scatter Plot (Figure 16) shows that 5Year periods of above 3.5% Real GDP Growth coincided with double digit equity returns for all 3 countries, whilst Real GDP Growth averaging below 1%pa produced a couple of single digit annualised equity returns. Real GDP Growth between 1%pa and 3.5% pa … no obvious pattern.

Of course, this analysis is a little simplistic as it starts at a fairly random start date, i.e. when the equity returns data commence, and hasn’t explored beyond this one possibility.

Source: Delta Research & Advisory

It is important to point out that the analysis in this paper is simplistic insofar that it only looks at quarterly versus quarterly, annual versus annual, 5 yearly versus 5 yearly and considers quarterly and annual lags in equity returns to see if the sharemarket is likely to be a leading indicator. There are many other timeframe permutations that could be tested and the analyses presented in this paper provides no indication as to the true impact of economic news or sharemarket news or the ability to predict stronger or weaker economic outcomes will have on equity market return predictions.

Most importantly, **there does appear to be a relationship between Real GDP Growth and Equity returns**. There is, however, a big question mark as to what that relationship is because what has been shown in this paper is that the relationship differs between country and across different timeframes and lags.

The sharemarket might be a leading indicator for future economic growth and the ability to predict economic future growth might help predict future equity market returns. But, there are still very large risks of failure involved. At the risk of stating the obvious, there are many other factors to consider, and whilst this paper doesn’t address or compare, valuation metrics are still likely to be a critical factor in the future equity return expectation. That said, whilst the ability to predict future economic growth might not always help in predicting equity market success, it still may provide an edge from time to time.

]]>Continue reading »]]>

]]>

Continue reading »]]>

When disappointing performance occurs, alarm bells will typically ring in the minds of investors, advisers, asset consultants and perhaps the managers themselves. Investing has only ever been a long game but thanks to the internet, the 24-hour news cycle, social media, etc. etc., it appears that success is expected to occur quickly and this is the case with investing too. Investors have such an enormous selection menu and if one strategy doesn’t appear to be working out then it’s not difficult to find another … and this has been exacerbated with the reduction of transaction costs.

Investment performance has 2 major problems:

- Analysis timeframes are too short
- Only performance (or benchmark relative performance) is observed

Pure investment performance (or benchmark-relative performance) on a day by day, month by month and even year by year basis is mostly noise and therefore a somewhat redundant analysis exercise. Different styles, risks, industries, sectors, and markets, constantly go in and out of favour and the analysis of performance over these short timeframes is typically a waste of time that can lead to high transaction costs and potentially even worse performance.

The purpose of this article is a simple one. It takes a back-to-basics approach of performance analysis and demonstrates, using a simple case study, one approach of how to cut through the noise to find the signal.

The signal defines what is really happening. The signal is the bigger picture investment view that enables us to undertake better investment analysis, and therefore construct better investment portfolios.

The goal of our case study is to analyse two actively managed Australian equity strategies, renamed to Blue Fund and Orange Fund, to determine their respective suitability for an investment portfolio.

Some quick simple statistics of each fund to begin with…average monthly return and volatility (i.e. standard deviation) over the same time period used in Chart 1 shows the best performer is the Blue Fund, followed by the benchmark, and then the Orange Fund. Low volatility is typically preferred and despite performance volatility has the same order, somewhat counter to the old adage that higher return requires higher risk.

Based on this simple return analysis many would already say that the Blue fund is the superior fund based on the higher level of return per unit of volatility (similar to Sharpe Ratio). However, it is essential to know that Table 1 also suggests that the average monthly returns are __not__ statistically different between both funds and MSCI Australia GR due to the high levels of volatility.

Source: Delta Research & Advisory; MSCI

This statistical lack of difference between returns is supported by Chart 1 which shows the monthly excess returns of each strategy compared to the MSCI Australia GR index. There is no pattern whatsoever and no one data point is likely t be predictable of the future. Sometimes the Blue fund outperforms the Orange fund and vice versa, and they both outperform the index at different times and vice versa. Thanks to the short time period used for this analysis, i.e. Monthly, this chart provides little to nothing and is a very good demonstration of noise with no obvious sign of a signal.

Source: Delta Research & Advisory

The next step in this analysis progression is to increase the period of analysis for each time series data point from monthly to 6 monthly rolling and then 3 year rolling periods. Six months is a common time period used between client portfolio reviews and Chart 2 shows substantial volatility of these 6 monthly excess returns and therefore reasonable evidence alone that any one six-monthly period should not be used to make an investment decision based on performance…and I’m sure that never happens. Six monthly rolling average appears to show no signal for these two funds individually, so the search for a meaningful signal continues.

Source: Delta Research & Advisory

On the other hand, potentially there is a signal from Chart 2 which includes timing differences between the Orange and Blue fund’s excess returns. That is, there does appear to be signs where their respective excess returns are moving in opposite directions, plus there are potential signs of the Orange Fund showing more extreme levels of outperformance or underperformance … however, at this point in the analysis these extreme levels should probably be taken with a grain of salt as it has only happened a few times … may be just good or bad luck???

So, moving from monthly returns to rolling 6 monthly rolling returns (Annualised), has yielded some potential insights but nothing conclusive.

Chart 3, takes the moving average of excess returns out to 3 years and the signals are getting a little stronger. The Orange fund has been a consistent underperformer on rolling 3 year periods since the middle of this time frame.

Anecdotally, 3 years is a popular timeframe for analysis and appears to be the amount of time investors and advisers are prepared to sustain underperformance of any one strategy before changing. The fact the Orange fund has had sustained underperformance over much longer than 3 years suggests that many investors or advisers may have excluded this fund from their consideration set or removed it from portfolios altogether.

However, the Blue Fund is looking very impressive, given very consistent rolling 3-year outperformance through almost the total timeframe so is looking to be the better strategy … or is it?

Source: Delta Research & Advisory

Risk-adjusted performance is frequently performed but rightly or wrongly it is rarely done on rolling timeframes and is usually performed across a single chosen time period. Unfortunately a single time period reduces a significant amount of information about investment performance behaviour so the risk-adjusted analysis performed here continues to to be time series based.

The Capital Asset Pricing Model

has a relatively poor reputation for predicting future returns. However, it is an excellent method for calculating exposure to the market (i.e. Beta) and risk-adjusted added value (i.e. Alpha). As we know, all active strategies aim to prove themselves with positive Alpha, which often doubles as the measure that defines “skill”. Using Capital Asset Pricing Model, performance analysis of both funds yields a market-risk-adjusted Alpha over rolling 3 year periods in Chart 4.

Chart 4 shows very different behaviours and potential signals for each fund. Firstly, their respective trends are in relative opposite directions, potentially suggesting opposite style.

The Blue fund has become negative in the earlier years, which wasn’t the case when analysing excess returns so there some currently unknown reasons for its market outperformance, as shown in Chart 3.

The Orange fund continues to show relatively poor results in the latter half of the timeframe, so its still looking like using this analysis many investors may have sold out of the Orange fund given this apparent weak performance.

Source: Delta Research & Advisory

The final risk-based performance analysis takes the Capital Asset Pricing Model a step further and introduces other systematic risks into the mathematical model…this model is based on the multi-factor model sometimes called the Arbitrage Pricing Model.

Possibly the most well-known of the multi-factor models is the Fama-French 3 Factor model which adds two risk factors, value and size, to the single risk factor Capital Asset Pricing Model. For the purposes of this analysis, there are 3 additional risk factors which combine MSCI defined indices; Value minus Growth, Small Cap minus Large Cap, and a Momentum risk premium to MSCI Australia benchmark.

Chart 5 shows the Alpha that remains after adjusting for all four risk factors and we can see a dramatic change in result for the Orange Fund. It now has positive Alpha for almost all of the time period analysed, whereas the Blue Fund has a multi-year period of negative Alpha after adjusting for these multiple systematic risks. So whilst the Orange Fund produced underperformance compared to the market in the latter half (refer Chart 3 and/or 4) of this time period, after adjusting for common equity market risk factors, it’s added value (Alpha) is positive. This positive Alpha is what we would want to see from an active manager; positive value-add over and above potentially cheap and replicable systematic risks.

Source: Delta Research & Advisory

So introducing non-market systematic risks improved the Alpha of the Orange Fund and slightly reduced the Alpha of the Blue Fund…how did this happen? The answer is a logical one…one or more of the systematic risks had a significant contribution to the performance of the Orange Fund…and the answer is shown in Chart 6. These two funds have clear distinct styles…the Blue Fund is clearly a Value-style (which MSCI define as holding low PE, low Price/Book, and/or High Dividend securities) and the Orange Fund clearly has a Growth style (which MSCI define as holding stocks with high earnings growth, revenue growth, and internal growth).

So whilst a fund manager may define how they invest with respect to style, deeper performance analysis can show whether that fund does what it says (i.e. is true to label), or can uncover what other risk exposures may be driving good or bad performance. It is worth noting that some qualitative research reports do not specifically define the Orange Fund as a “Growth” fund but given the track record, and the definitions of Growth, it is difficult to argue with the performance analysis.

Source: Delta Research & Advisory

The process outlined above will differ from strategy to strategy and will depend on what is important with respect to the role of a potential strategy within a broader portfolio. However, hopefully what this process shows is the significant benefit in undertaking performance analysis that looks through the short term noise that all strategies incur and look for signals using longer timeframes.

To find the signal for constructing better investment portfolios also requires digging deeper into a strategy’s return series. It may involve adjusting performance for multiple risks, such as the market or styles (Value, Growth, Size, Momentum, Quality, Credit, Duration, etc.) whilst still increasing the time period of analysis. Moving away from quarterly or annual performance measures that dominate the industry and survey data is essential. Moving away from analysing performance over short time periods is a positive move that in aggregate will result in superior portfolio performance from lower transaction costs.

]]>The above chart shows the yields for Australian Government Bonds, both nominal bonds and indexed bonds, as at the end of last week (although you can adjust the pricing date to any trading day of 2016). A simple way to determine the market’s inflation expectations over different timeframes is to simply subtract the difference. If you look hover the mouse over each line chart around the maturity date of circa 2020, the difference in yields is in the vicinity of 1.2% to 1.3% … low inflation expectations indeed and certainly a lot lower than the RBA target of 2% to 3%.

So if the RBA achieves its target, then these Australian Government indexed bonds will prove to be pretty reasonable performers … or at least relative to their nominal counterparts. I do know that most of the financial planning industry are still using 2.5% as their inflation target, or at least between 2% and 3% … clearly the market is currently thinking this is way too high and lower inflation than most of us expect should be expected.

PS … it also means the RBA cash rate is very likely to decrease.

]]>Continue reading »]]>

Over recent years many commentators and experts have spoken of the significant risks superannuation funds are carrying with respect exposure to Australian equities. Most notable were comments a few years ago from David Murray, former Chairman of the Future Fund, and Ken Henry, former Federal Treasurer, who both said they had concerns that Australian superannuation funds were overweight Australian equities. Probably the comment I heard the most, coming from many investment professionals, was along the lines of, “balanced funds have around 60% to 70% in equities but this accounted for more than 90% of the portfolio risk”. So I thought I’d finally get around to checking out how true this statement is and if the industry as a whole has changed much over the last few years with respect to the influence of equities and their multi-asset portfolios.

If you trust my analysis and aren’t deeply familiar or interested in quantitative methods, then save yourself some sleepy time and skip to the results. Otherwise …

There are two primary analyses undertaken, and both involve regression analysis. Instead of analysing individual diversified strategies, I have chosen the following peer group indices as they pretty much capture the market as best as I can find …

- Morningstar Australia OE Multisector
**Conservative**(0-20% Growth Assets) - Morningstar Australia OE Multisector
**Moderate**(20%-40% Growth Assets) - Morningstar Australia OE Multisector
**Balanced**(40%-60% Growth Assets) - Morningstar Australia OE Multisector
**Growth**(60%-80% Growth Assets) - Morningstar Australia OE Multisector
**High Growth**(80%-100% Growth Assets)

These indices are also chosen as they pre-tax, thereby, producing an apples and apples comparison with the benchmarks which are also pre-tax.

The following benchmarks represent Australian equities, Global equities, and the risk-free rate …

- MSCI Australia GR AUD
- MSCI World GR AUD
- Bloomberg AusBond Bank 0+Y TR AUD

The following models were used to assess contribution to portfolio risk by equities and the portfolio exposure…

- R
_{p}-R_{f}= α + β_{1}(R_{a}-R_{f}) + ε - R
_{p}-R_{f}= α + β_{1}(R_{a}-R_{f}) + β_{2}(R_{w}-R_{a}) + ε

Model 1 is used to calculate exposure and contribution to total portfolio risk by Australian equities and Model 2 is used to calculate the exposure and total contribution to risk by both Australian equities and global equites.

These regression models are applied to monthly returns between 31/12/1993 and 30/4/2016…which is a long time!

The variables from the models are:

- R
_{p }is the monthly return of peer group index - R
_{f}is the Risk-free rate - α is the Alpha of the model (or beta-adjusted excess return)
- β
_{1}is the calculated exposure to Australian equities - R
_{a }is the monthly return of Australian equities - β
_{2}is the calculated exposure to the excess return of Global equities minus Australian equities - R
_{w}is the monthly return of Global equities - ε is the residual error of the model

The R-squared value of each regression equation is calculated to determine the portfolio risk that can be explained by each model and is therefore used as a proxy for “risk contribution”. The R-squared of a regression model is also known as the “goodness of fit” and its calculation (without going into too much detail) is = “Explained Variation”/”Total Variation”.

**Chart 1 – Australian Equity Beta**

The above chart shows the Beta, or exposure, of Australian equities to each of the peer group indices. As expected, the higher the allocation to growth assets, the higher the exposure to Australian equities market. Interestingly, since late 2011 it appears the Australian equities beta has declined suggesting a lower exposure. Given the maximum growth assets for each peer groups isn’t much higher than the Australian equities beta for each peer group, you could interpret that Australian equities is the dominant asset class, and maybe it is. But, it may also suggest it is evidence of the relative high correlation between Australian equities and other growth asset classes.

Either way, these results are consistent with expectations and may somewhat support concerns around higher Australian equities allocations given their exposures or sensitivity appears to be a high proportion, but it also proof that this sensitivity to Australian equities has been in decline over the last few years or so.

**Chart 2 – Australian Equity Contribution to Portfolio Risk**

Now whilst the exposure to Australian equities appears fairly consistent with expectations; its total contribution to portfolio risk is a different story. In essence for the most part over the last 23 years (this chart starts the end of 1996 and is rolling 3 years so really starts in 1993); the Australian sharemarket contributes to a majority of risk across all multi-asset class peer groups and is therefore a very very important part of the portfolio.

Over the last few years, which is the very last point on the far right of the chart, this percentage is in the vicinity of 63% to 83% across each peer group which is somewhat consistent with the concerns spoken of Balanced funds by various experts … that is, “60% allocation to growth assets but responsible for 90% of the risk”. However, even for conservative strategies where the allocation to growth assets is less than 20%, Australian equities contributed at least two-thirds of the total portfolio risk over the last 10 years.

**Chart 3 – Australian + Global Equity Contribution to Portfolio Risk**

Chart 3 shows the same risk contribution statistic as Chart 2, but this time it is for Model 2, which adds the global equity market. The increase in risk contribution from both equity markets is only marginal because we are adding only the excess return by global equities over Australian equities and the two markets are fairly positively correlated so the impact of the additional asset class is small.

The more interesting results from Chart 3 include, over the last 10 to 15 years Australian Equities and Global Equities account for…

- More than 90% of the total portfolio risk across Balanced, Growth, and High Growth peers of multi-asset class strategies.
- More than 80% of the total portfolio risk for Moderate peers of multi-asset class strategies
- Between 60% and 80% of total portfolio risk for Conservative peers of multi-asset class strategies, despite no more than 20% allocated to growth asset classes!!!

So irrespective of the allocation to equities or the equities market beta, across all risk profiles, equities are clearly the dominant asset class in terms of contribution to total portfolio risk.

So when “experts” say that “equities account for 90% of total portfolio risk of a balanced fund” therefore implying there is too much exposure, it is not necessarily about too much exposure just the importance of equities. So what should investors do to reduce this reliance on equities? As we see above, only holding 20% maximum of growth assets like the conservative peer group, still produces a very high proportion of portfolio risk due to equities.

The answer is to include non-correlated assets…or in English, add assets to the investment portfolio that behave differently from equities and go up when equities go down. This reverts to Markowitz 101 and is the continued search for the holy grail of investing…including non-correlated assets to the portfolio that can reduce the risk without reducing the return expectation or increase the return without increasing the risk.

The obvious non-correlated asset over many years has been conservative highly rated bonds…which I believe was Ken Henry’s suggestion when looking to reduce the reliance on equities. Adding conservative bonds to a portfolio did reduce the contribution to risk from equities (see Charts 2 and 3 above) but as we know, adding conservative bonds is unlikely to improve the return expectations from equities and by our industry’s definition, obviously changes the risk profile. A good example of reducing equity risk are lifecycle funds. These have a moving risk profile (i.e. decreasing through time) and they gradually increase a fund’s exposure to bonds throughout time. The effect of this is to reduce the size of the volatility to combat sequencing risk leading into and through retirement, but the volatility will still be most dependent on equities.

Other potential lowly correlated considerations are alternatives, like property, private equity, infrastructure, or perhaps hedge fund strategies. There is much debate about the value of some alternatives (asset consultants and fund managers in favour of Alternatives and some big institutions are throwing in the towel, i.e. CALPERS), and if you do believe alternatives are the diversification solution, significant care must be taken to truly understand what is driving the underlying risk of these strategies. Particularly because equity markets may still be a very influential driving factor!!! Either way, if alternatives do reduce the reliance on equities in portfolio risk equation, they do so by introducing other risks … which is not necessarily a bad thing but may be. So the challenge then becomes about assessing whether those risks are adequately compensated.

The investor faces very challenging times. Interest rates both here and around the world are so low, that retired millionaires are at significant risk of running out of money. To produce higher returns still requires the acceptance of higher risks but escaping equity market risk is not at all easily achieved without significant sacrifices in costs, liquidity, or chancing the unknown. So no matter what the investor’s investment strategy or risk profile, there will most likely be a strong reliance on equities driving their success. So please note, the communication of this bigger picture concept will always be more important than the marginal advantages gained or lost from manager selection, dynamic asset allocation, security selection or whatever the latest trend is.

]]>Continue reading »]]>

When choosing strategies or managers for each asset class, portfolio constructors will often look at capturing or avoiding particular systematic risks. For example, we may combine “value” and “growth” styles or perhaps acknowledge particular anomalies and bias a portfolio towards a “value” or “quality” styles across a variety of equity markets, and other systematic risks such as size, momentum, low volatility, or perhaps illiquidity may also be considered, amongst others. However, I think it’s fair to say the impact of systematic risks is rarely considered across asset classes … so I believe there may be a few unanswered questions of which one is … what systematic risks does the Australian equities asset class bring to the global equities asset class?

I’m glad I asked…

Without a deep dive analysis, one might argue that compared to major developed markets, Australia’s high dividend yield, currently lower PE Ratio, and generally smaller companies, one might think the Australian equity market behaves like a global small cap with a value style tilt. Perhaps the high commodities exposure and economic link with China might suggest there’s a growth component and/or behaviour that is emerging markets-like. So what is the truth?

**Chart 1**

Source: Delta Research & Advisory

Without boring you with too much detail with respect to the analysis leading up to the above chart. I can tell you that Chart 1 shows the risk contribution or risk “make-up” of the Australian equity market based on a number of global equity market risk factors. Namely,

- Global Market Risk (MSCI World GR AUD risk premium to Cash)
- Value (MSCI World Value minus MSCI World Growth)
- Size (MSCI World Small minus MSCI World Large)
- Emerging Markets (MSCI EM minus MSCI World GR)
- US Dollar (US Dollar vs Australia Dollar)
- Idiosyncratic (Everything thing else) … which I assume may be mostly Australian company specific risks

Each point on the chart represents a rolling 3 year contribution to risk of the various factors mentioned above. Whilst numerous studies of managed funds have shown high levels of risk contribution from the market (e.g. more than 90% from the famous Brinson, Hood, Beebower study of US Pension funds), the contribution to the Australian equity market risk from the global equity market has been consistently less than 50% since the start of this analysis in 2001 (don’t forget the rolling 3 years, hence why the chart starts at 2004).

The global Value factor (Purple) barely gets a look-in, but there are strong risk contributions from…

- Size (Red colour – until the 3 years to mid-2014)
- Emerging Markets (Orange), and
- US Dollar (Green – particularly since the 3 years to mid-2007)

Idiosyncratic risks (or those risks that can’t be captured by the others – brown colour) are a very large contributor to total risk and over the last three to four year has contributed between 60% to 80% of total Australian equity market risk. From a portfolio construction perspective, this may be a good thing. Why? Because, these idiosyncratic risks, which are possibly mostly Australian company-specific, are uncorrelated with global equities and the other systematic factors…this means the Australian equities asset class does carry with it a reasonable level of diversification…or at least lack of correlation.

So, having established the contributions to risk, how big are these contributions?

Chart 2 shows the global market beta has averaged around 0.55 over the last 15 years or so, suggesting the Australian market in the context of global equities is a low beta strategy. As shown in Chart 1, this risk factor is not the majority contributor to risk, so there’s a lot more to Australian equities than behaving like a low beta Global equities strategy, but nevertheless it is still significant and having lower relative risk than the global share market may not be a bad thing.

**Chart 2**

Source: Delta Research & Advisory

Given its low contribution to overall risk, unsurprisingly, the Value risk factor is pretty much zero…it jumps between being “value”-like (above the zero line) and “growth”-like (below the zero line) to average slightly growth-like at -0.08 but it’s not at all statistically significant so can be ignored. The Australian market is neither a value or growth market.

The Size factor, has mostly been positive and strongly so. This suggests the Australian equity market has mostly behaved like a small-cap biased global equities strategy … and considering the average market cap of the Australian market that should also come as no surprise given what we know…let’s face it, Australian companies are nowhere near as big as Apple, Microsoft, etc. In fact, the combined market cap of those 2 companies alone is more than the Australian equity market as a whole.

The Emerging Markets risk factor, which I’ve treated as an extension to the Global Market, is also significantly positive suggesting that Australian equity market performance has a positive correlation with the performance of Emerging Markets. However, with levels consistently around 0.5, the Australian market also has lower Emerging Markets beta so has slightly lower price volatility impact. This is probably due to our close trading relationship with China coming through, but at least it comes with slightly lower risk.

And the final systematic risk, being the US Dollar relationship with the Aussie dollar, is also a very significant factor utility of over 0.5 (and up to 0.8 in recent years) suggesting that on average, when the Australian dollar falls by 10%, it has a performance drag on the Australian market by around 6%. This US Dollar factor might be regarded as an economic factor, as the Australian dollar weakens when the economic outlook weakens, so this result is also not too surprising.

That leaves, the “Pure Alpha” which is the additional return the Australian equities market provides after the above-mentioned systematic factors are removed. Given the significant overweight most portfolios have to Australian equities, it is pleasing to say that since 2001 this higher allocation has produced an average risk-adjusted alpha of a positive 2.9%pa. The alpha hasn’t always been positive, particularly in recovery years following the GFC, but it does suggest the inclusion of Australian equities as an asset class has been a worthy one … and these figures don’t include franking credits which would add even more alpha (say around another ~1.5%pa).

So all-in-all, over the past 15 years the Australian equity market has shown it is driven by some global systematic factors including global equity markets, emerging markets, global small caps, plus some US Dollar influence. The Australian equity market cannot be seen as either value or growth and a large proportion of its performance is probably unique to Australian conditions and hopefully the positive alpha that has been generated over that time can continue.

]]>Continue reading »]]>

The goal of this article is to set a baseline for Australian equity market return expectations using a long(ish) look at a few key measures. Now, I do realise there are very few believers that past prices can tell you something about the future…but…and you knew that but was coming…there is a somewhat remarkable consistency to the trend of equity market returns over the last 45 years as shown in Chart 1.

**Chart 1 **

Source: MSCI, Delta Research & Advisory

Sure there is a lot of volatility but looking through all of that, since the end of 1969 (and that’s all the data I have), the trend-line through Australian equity returns is flat…that is…it is pretty much same as the average return. Now when it comes to measuring performance, in my opinion and many others, using a trend line is a superior method to simple point-to-point return analysis as it reduces starting and ending point biases. A second observation of Chart 1 and this flat return trend and maybe in spite of this trend, is that equity market volatility appears much higher during the high 1970s and 1980s then afterwards. So we have varying volatilities but a flat trend line of returns.

Now one of the numerous factors driving equity market success is the state of the economy. Chart 2 shows GDP (seasonally adjusted) over the same time period, i.e. since the end of 1969, and whilst direct correlations between equity market returns and economic growth may be weak, Chart 2 does suggest two similarities. Firstly, the trend line, like with equity markets, is flat over the 45+ year period, and once again, the 1970s and 1980s are also more volatile than the following period.

**Chart 2**

Source: ABS, Delta Research & Advisory

Now whilst GDP and equity market returns have shown flat trends (albeit with volatile volatility), there is one financial market that has changed significantly … interest rates. Chart 3 show quarterly returns of the RBA Bank Bill 90 day index since 1969 and this time there is no sign of a flat trend and little doubt the trend is down … and significantly so…but this should come as no surprise. The high inflation 1970s and 1980s resulted in high interest rates whilst over the last 25 years, has seen the lower inflation provide the Reserve Bank with scope to reduce rates to deal with the ongoing maintenance of economic growth and employment. Corresponding with this, is once again the same volatility pattern…high volatility in the 1970s and 1980s and since then…a significant reduction since.

Now when forecasting long run market returns, valuation is typically a very important factor along with many other metrics (depending on who you ask) and I don’t intend to address those here. As mentioned, the intention of this short piece is to set a baseline for expectations that can be expended and improved.

The above trend analysis suggests a real possibility that average equity market returns may continue to trend sideways …certainly 45 years (as much data as I have) isn’t a bad start. I believe the main question around what equity markets will do starts with the future trend of inflation…which means the Reserve Bank plays a key role as it’s key operating philosophy is that it believe it can contribute to Australian economic prosperity by setting the cash rate to meet an agreed inflation target.

So if inflation stays low, so too will interest rates, and déjà vu, we may continue to see what has more recently been happening. If on the other hand we experience higher inflation, interest rates will increase, which may increase equity market volatility as the denominator of valuations are placed under pressure.

**Chart 3**

Source: RBA, Delta Research & Advisory

There is, however, one inflationary scenario not captured in the Australian economy and financial markets since 1969 and is rarely talked about; at least in the Australian context…deflation. There is definitely a complacency in the Australian economy which is probably due to the fact it has been so long since Australia actually had a technical recession. Whilst deflation is not a high probability scenario, Australia does face numerous challenges that don’t rule it out altogether. Australia’s economy is transitioning to new economic drivers; the resources investment boom is over; population is ageing; Australia has amongst the highest levels of household debt in the world; and this is accompanied by a residential property market that many argue is in a bubble in the largest markets of Sydney and Melbourne.

A severe bursting of this bubble accompanied by deflation (as seen in Japan, Spain and other burst property bubbles) would mean the above scenarios of sideways trends in equity markets and economic growth do not apply. Interest rates don’t have far to fall before reaching the zero bound and the only market mechanism left would be further weakening of the Australian dollar.

So the simple baseline scenario for long run return expectations to build from is …

- Continued contained inflation = continuation of the long run trend in equity markets and similar volatility to last 25 years + continued low interest rates
- Higher inflation = continuation of long run trend accompanied by much higher volatility then currently used to + much higher interest rates, or
- Deflation = sharp decline in the current trend accompanied by high volatility + hitting the zero interest rate bound

…or something completely unprecedented.

PS…email me if you require further proof of the above trend analysis

]]>Continue reading »]]>
**Background**

There’s a widely held belief that to create alpha (i.e. positive returns after adjusting for risk…let’s say market risk), a manager needs to make meaningful bets away from the market. That is, stop being a “benchmark hugger”, concentrate the portfolio with best ideas, and/or move the portfolio holdings away from the benchmark and possibly be more absolute-return oriented. We have all seen numerous strategies that meet these criteria, have generated strong alpha, but is this a reality or it is a belief that is lacking in evidence, save a handful of strategies that just so happen to tell us so. This article seeks to provide some clues to whether accepting greater non-market risk does produce higher alpha.

Firstly, let’s define non-market risk?

Possibly the most frequently used measure is tracking error (which is the standard deviation of the difference between a portfolio’s returns and its benchmark). Whilst tracking error is a good measure of non-market risk, it can be a little misleading in the example of a geared index fund as there is absolute are no bets away from the market due to being an index fund, but the gearing produces a high tracking error. Therefore I believe tracking error creates a potentially inaccurate bias when it comes to comparing non-market risk to alpha generation.

A more recent popular statistic is active share which describes the percentage of holdings that are different to a market benchmark. A portfolio with a high active share suggests a large difference from the benchmark. Because this statistic is a holdings based measure it is quite difficult to measure on a regular basis. Secondly it is also possible to have a portfolio with a high active share but is highly correlated with the market benchmark suggesting that holdings differences do not necessarily translate into performance differences.

A preferred measure of non-market risk is what some may also call idiosyncratic risk. Idiosyncratic risk is similar to tracking error but is adjusted for exposure to market risk (i.e. market beta) and is defined as the proportion of a total portfolio’s risk due to non-market bets. To be specific, it is (1-R²), where R² is the goodness of fit of the Capital Asset Pricing Model (CAPM) to the portfolio in question … CAPM is represented by Equation 1 below. A second advantage of using idiosyncratic risk is that Equation 1 is also used to calculate Alpha…so a win-win.

So the statistic this paper places most emphasis on is… α/(1-R²)

…and after checking numerous textbooks, I cannot find a name for it, so for this paper will declare it the “Furey Ratio” until someone corrects me. The Furey Ratio is similar to the Information Ratio (which is the ratio of excess benchmark return divided by the portfolio tracking error) but unlike the Information Ratio, the Furey Ratio adjusts for different levels of market risk. So the Furey Ratio is another measure of risk-adjusted return and is Alpha per unit of Idiosyncratic Risk. What we really want to see in an active manager is a high Furey Ratio, meaning they are getting big bang for their non-market risky buck!

The analysis plan is to assess whether managers are more likely to produce higher risk-adjusted alpha with greater idiosyncratic risk … so to do this we will test the statistical significance of the Furey ratio, α/(1-R^{2}), from a sample of manager returns.

The manager returns uses performance from two groups of strategies…

- Global Equities (Sample size = 121)
- Australian Equities (Sample size = 226)

…which are the two largest equity asset classes in the Australian investment landscape.

Monthly performance from September 2010 to September 2015 is used and acquired from Morningstar Direct. Duplicated strategies, where the only difference is fee structure, are removed.

The 5 year time-frame to the end of Sep 2015 has been chosen for the following reasons…

- 5 years produces sufficient numbers of both monthly performance (i.e. 60) and number of available strategies
- It is after the Global Financial Crisis period of 2008/09, i.e. from Sep 2010
- It balances the survivorship bias that comes with using a longer time-frame with a reasonable overall sample size…i.e. survivorship bias is a real issue if considering the GFC period as only the better managers survived through to 2015 from before the GFC period

Aside from these reasons, 5 years is still a somewhat arbitrary time period (i.e. it probably makes little difference compared to 5 years and 2 months of data).

Chart 1 shows CAPM Alpha vs CAPM Idiosyncratic risk over 5 years to end of September 2015 for the 121 Global Equity Managers taken from Morningstar Direct database. All managers chosen have a minimum 5 year track record, and are classified by Morningstar as Global Equities managers.

On the positive side for global equity active managers the regression line in chart 1 slopes upwards suggesting there is a chance that with greater non-benchmark risk comes from higher alpha (CAPM Alpha). This trend demonstrates a positive Furey Ratio but unfortunately, the P-value of 0.153793 of the trend line suggests it is not significantly different from zero at the usual required minimum significance levels (i.e. 0.05)… so weak evidence that higher alpha is not strongly correlated with greater non-benchmark risks.

Source: Delta Research & Advisory

A simple observation from Chart 1 is that there is significant clustering of values at the lower end of the x-axis and a fanning out of alpha levels as Idiosyncratic risk increases. This suggests there may be a reasonable argument that regression analysis of this data may be somewhat inappropriate. So to counteract this issue, the following analysis divides the above CAPM Idiosyncratic risk measure into 5 quintiles.

Source: Delta Research & Advisory

Once again, there are positive signs as there is higher Alpha for the two higher quintiles of Idiosyncratic Risk. However, and unfortunately for active managers, the higher values are not statistically different…please refer the following Hypothesis Test between quintiles 3 and 5 (which have the largest difference).

Source: Delta Research & Advisory

So stopping the analysis of Global Equities strategies there, so far there is little evidence to suggest a statistically significant and positive Furey Ratio amongst the Australian market of Global Equities managers over the last 5 years … therefore suggesting higher idiosyncratic risk probably hasn’t produced higher alpha.

Similar to Global Equities strategies chosen, a sample of Australian Equity managers have been chosen from the Morningstar Direct database, duplicated strategies have been eliminated, and 5 years of monthly performance between September 2010 and September 2015 used for the following analysis.

Chart 3 shows that once again there is the spread of Alpha as Idiosyncratic risk increases and the slope of the line (i.e. Furey Ratio) increases. Also similarly, the Furey Ratio is not significantly different from zero at the 5% level (P-value = 0.07215 which is greater than 0.05), also indicating this chart does not suggest greater alpha from higher levels of idiosyncratic risk…at least using statistical tests.

Source: Delta Research & Advisory

Like Global Equities there is a reasonable argument that the regression analysis is not appropriate due to the larger variance of Alpha as Idiosyncratic risk increases so similar group analysis is applied by dividing Idiosyncratic Risk into quintiles and the results are shown in Chart 4.

This time there is a statistically significant difference between the Alpha of those managers at the 4^{th} quintile and those at both the first and second quintile…but not between the others (you’ll have to trust me on this)…please refer following Hypothesis Testing results.

Source: Delta Research & Advisory

Source: Delta Research & Advisory

Source: Delta Research & Advisory

Observing Chart 4, it does appear to show two distinct groups where quintiles 1 and 2 have Alpha results around 0 whilst quintiles 3 to 5 have CAPM Alpha of more than 1% on average…which I’m sure many active managers would be pleased about. Combining the quintiles into these 2 groups yields the following results for CAPM Alpha…

Source: Delta Research & Advisory

…and the difference in means are statistically significant given the Hypothesis rejection below…

Source: Delta Research & Advisory

So, if I may say that after some potential data mining, there may be some evidence that greater idiosyncratic risk relates to higher levels of alpha (or at the risk of being egotistical, a higher Furey Ratio) among Australian equities managers.

For those interested, the level of Idiosyncratic Risk that intercepts between Quintiles 2 and 3 is only 5.82% (which is around the borderline of the clustering in Chart 3)… meaning if the market, as defined by MSCI Australia GR, explains more than 94.18% (i.e. 1 – 0.0582) of an Australian equity manager’s performance volatility, then this may decrease the chances of generating positive alpha and vice versa.

Over the 5 years to September 2015, the evidence within this paper is possibly weaker than many would expect and shows there is little to no relationship between managers generating alpha and idiosyncratic risk…particularly for Global Equities strategies.

There is some evidence that greater idiosyncratic risk has led to higher alpha amongst Australian Equities strategies although it does __not__ appear to be a linear relationship. However, the result over the last 5 years does show that Australian equities managers have, on average, produced a significantly higher alpha if their non-benchmark risk is greater than around 5.8%.

So for Australian equities managers, the optimistic conclusion (so far) is that there are two groups…the first group is the much-maligned benchmark huggers (with idiosyncratic risk less than 5.8%) who have struggled to produce any alpha at all on average; and the second group with idiosyncratic risk that is higher than 5.8% which has produced a significantly higher alpha of 1.7%pa over the 5 years to September 2015. This result doesn’t mean the higher the idiosyncratic risk the higher the alpha (because of the lack of linear relationship) but it is some evidence that a higher non-benchmark risk does increase the chances of positive alpha…__so the jury is still out__ but benchmark huggers should beware.