There are 278 managed funds listed on the CommSec website. Every so often, I look at the pool to see how EGP is performing against the alternative options into which our investors might deploy their investable funds.
Usually when I make the comparison, I narrow the pool down to Australian equity focused funds, for there are a huge variety of different offerings available, examples:
· Multisector Aggressive
· Equity Global Resources
· Equity Australia Large Value
· Equity World – Currency Hedged
· Equity Australia Large Blend
· Equity Australia Large Growth
· Equity Global Real Estate
· Equity Australia Real Estate
· Bonds Australia
· Bonds Global
· Commodities and Precious Metals
· Equity Asia Pacific w/o Japan
· Hedge Funds Australia
· Hedge Funds Global
We have limited our fund to Australia, but if I see something I like, I will happily reach across sectors. The above list is only 1/3 of the fund styles available. EGP sit in the 9th percentile of this group (.xls spreadsheet), based on the 12 months to August and in the 10th percentile in the 3 year category. Had I narrowed the pool to Australian focused funds not using leverage, we would be similar in the 1 year, but probably top 5% over 3 (the last 3 years have been kind to those with international allocations, which is why we’re a little further back in that pack).
I don’t point this out to be boastful, though I confess to being very pleased with the returns we have generated for our investors to date, but to point out that it doesn’t take wildly exceptional results to drift near the top of any group over time.
This is sort of how I envision things playing out for EGP over time, by generally being in the top half (preferably the top half of the top half…), over a very long time-frame, really top-notch results will be generated.
What I had actually intended to do with the post this week is demonstrate how hard it is and how long it takes for investment results to be distinguished between ‘success’ and ‘luck’.
The second tab of the spreadsheet attached above shows a ‘theoretical’ group of 100 funds over a 20 year period. I used the random number generator function of excel to generate returns that would approximately simulate a 20 year equities period. The average return across the period was 8%. In any given year, the average return was between -25.4% (year 17) and +31.6% (year 18). The average return over the 2000 fund-years (100 funds x 20 years) was 8%.
The best randomised return out of the 100 ‘funds’ was 17.1% annually (Fund 22) and the worst was an annual loss of 0.8% (Fund 52). The best and worst funds were 2.6 standard deviations (σ – sigmas) above and 2.2 sigmas below the mean, in a normally distributed data-set you’d expect about 2.5 sigmas each way to cover the field.
After generating the randomised data for the 100 funds, I added another fund, the ‘30th percentile fund’, which is consistently just a little better than average. The manager of the ‘30th percentile fund’ was on average about 14% behind the best performer in each year, and never really stood out from the crowd. After 10 years of being in the 30th percentile, the good performance was barely distinguishable from luck, although the ‘30th percentile fund’ was ranked second out of the 101 funds, there was still 5 other funds within 2% of the ‘30th percentile fund’s average annual return. Still, after 10 years of 30th percentile performance, the evidence was starting to point to skill rather than luck. But when you’re only 2 sigmas away from the mean after 10 years, you will have your doubters.
After 20 years, when the ‘30th percentile fund’ has an annual advantage over the next best fund of 1.7% and is now a little more than 3 sigmas better than the average, the explanation of luck starts to become untenable, for at 3.3 sigmas, there is less than 1 chance in 1000 that luck has been the driver of outperformance.
If you ever wonder why the professors were so keen (.pdf – Mungers 2003 UCSB speech) for such a long time to declare the results generated by Buffett/Berkshire as luck, this is why, because it takes so long to disprove. Buffett last time I heard it calculated was out past 7 sigmas, which makes it less than a one in 400 billion chance the record is just luck.
EGP has got our first sigma behind us, so we are is still trying to put our second sigma in the rear-view mirror at the moment, fortunately I have plenty of years ahead of me – Tony Hansen 29/08/2014
|
|
Apr 1st 2011 |
Jul 1st 2014 |
Current Price |
Current Period |
Since Inception |
|
EGP Fund No. 1 |
1.00000 |
1.56145 |
1.60474*1 |
2.77% |
75.01%*2 |
|
35632.05 |
45991.23 |
48315.28 |
5.05% |
35.60% |
EGP Fund No. 1 Pty Ltd. Up by 2.77%, trailing the benchmark by 2.28% since July 1st. Since inception, EGP Fund No. 1 Pty Ltd is Up by 75.01%, leading the benchmark by 39.41% all-time (April 1st 2011).
*1 after 31May 2013 dividend of 2.333 cents per share plus 1.000 cent per share Franking Credit & 31 May 2014 Dividend of 7.000 cents per share plus 3.000 cent per share Franking Credit
*2 calculated based on dividends reinvested
