The end of 2016 was an important milestone for the UK Value Investor model portfolio. It marked the end of the portfolio’s fifth full calendar year and so at last it is beginning to build up a meaningful performance track record.
I say “meaningful” because stock market returns over periods of less than five years are little more than random noise. In fact, I think an investment manager or methodology really needs a track record of ten years or more before you can say anything robust about its performance.
And so with five calendar years in the bag, the end of 2016 is a good point to look back and review the portfolio’s toddler years.
But first, here’s a very quick recap of how this model portfolio works:
- It started on 1st March 2011 with a virtual £50,000 (tracked using ShareScope)
- It’s benchmarked against another portfolio of the same size which holds a FTSE All-Share tracker (the Aberdeen UK Tracker Trust)
- All dividends are reinvested
- All expenses are taken into account, including broker fees (£10/trade), stamp duty (0.5%) and an annual subscription to the UK Value Investor newsletter (currently £270/year)
- It’s managed according to a defensive value investing strategy
- More than 90% of my personal savings are invested in exactly the same stocks
Goals: High growth, high yield, low risk, low cost
Before I review the portfolio’s performance I want to reiterate its four key goals. Clear goals are absolutely critical because if you aim at nothing, that’s what you’ll get. The portfolio’s goals are:
- High Growth: The portfolio’s total return over five years or more should be greater than the FTSE All-Share’s
- High Yield: The portfolio’s dividend yield should always be greater than the FTSE All-Share’s
- Low Risk: The portfolio’s maximum peak-to-trough decline over the last five years or more (also known as drawdown) should be smaller than the FTSE All-Share’s
- Low Cost: The cost of running the portfolio in terms of money (fees and taxes) and effort (time spent analysing companies) should be low (which I’ll define shortly)
This is a balanced set of goals. I want a high yield but not at the expense of growth, and I want high growth but not by taking on more risk than I would through a passive index tracker.
Here’s how the portfolio has performed, relative to those four goals:
High Growth: Beating the market by about as much as I’d hoped
My target for total return (capital gains plus dividend income) is to beat the market over periods of five years or more. This chart shows how the portfolio has done so far:
My primary goal was to beat the market and clearly that has been achieved.
My secondary goal was to beat the market by 3% on an annualised basis, and that has been achieved over five years but not quite achieved since inception in 2011.
However, I’m not massively worried as it’s still early days and the investment strategy is much more mature today (and I expect much better at producing results) than it was in 2011 and 2012.
What about returns in 2016 I hear you ask?
To be honest, I have gone off the idea of tracking one-year returns. I have always been against tracking short-term performance, but in the past just went along with what pretty much everyone else had always done, which is to talk about results over the last year (calendar or rolling).
But having recently read an excellent 2016 review and rant against tracking short-term performance, as well as The People’s Trust’s commitment to target returns over a seven-year period, I have decided to ditch what is essentially an unhealthy and pointless habit.
Most stock market investors know it’s daft to worry about whether their portfolio beat the market yesterday, last week or last month.
It’s daft because the market is noisy, and returns over shorter periods are driven by random fluctuations rather than any inherent strength in a company, portfolio or investment strategy.
The same logic against measuring returns over a day, week or month extends to measuring results over a single year.
Here’s an example: If a portfolio of 30 stocks has a 4% dividend yield and manages to grow that dividend by 6%, it would be reasonable to say it achieved a 10% return in that year. However, the market value of that portfolio could easily go up or down by 30% or more in that same year, simply because other investors decide to buy or sell those same stocks.
The market’s 30% random walk, driven by the buying and selling of other investors, completely trumps the 10% economic returns of the portfolio’s holdings.
However, over five years or more, those 10% returns will have accumulated to more than 50%, and by that point the economic returns of the portfolio’s holdings will have begun to exceed the affects of the market’s random ups and downs.
Or to put it another way: Short-term returns are driven by luck while long-term returns are driven by skill (or a lack of it).
That’s why equity investors are reminded time and again to only invest with money that they don’t need back for at least five years.
And so from now on I won’t talk about returns over the last week, month, year or even three years. Instead I’ll only measure returns over periods of five years or more, and I suggest other long-term equity investors, especially managers of collective funds, do the same.
High Yield: A consistently higher yield than the market
The goal here is simply to have a better yield than the All-Share at all times, and since inception the model portfolio’s has always achieved that, apart from a couple of forgiveable situations which I’ll explain in a moment. Here’s the yield chart:
The two periods where the portfolio’s yield was below the market’s were:
- The build up phase, where the model portfolio was built up during 2011/2012, and its dividends were correspondingly slow off the mark
- A brief period in 2015 where the FTSE All-Share tracker benchmark changed the percentage split between its annual and interim dividends. This temporarily skewed its yield upwards by around 1%
Neither of those periods reflects the true yield of either the model portfolio or its benchmark, so as far as I’m concerned the model portfolio has always met its goal of having a higher yield than the All-Share.
One other feature of the portfolio’s yield is that its gap over the All-Share’s yield has come down gradually over the years.
This happened because a couple of years ago I added profitability to my stock screen (measured as median ten-year return on capital employed, or ROCE) and the result is that I now tend to pick higher profitability companies than in previous years.
These higher profitability companies tend to be higher quality companies, so they tend to trade on higher multiples and with lower yields, and that’s why the portfolio’s yield has gradually declined.
The flip side is that I expect these higher quality companies to generate faster and more reliable growth in the future, so this is effectively a swap from higher yield stocks to higher growth stocks. I also expect the total return from these higher quality companies to be better overall.
Here’s another chart showing total dividends on an annual basis:
These annual dividend payments have grown slightly slower than the capital value of both portfolios. Measuring dividend growth from 2012 to 2016 gives the following results:
- Model portfolio dividend growth = 7.1% annualised
- FTSE All-Share dividend growth = 9.8% annualised
I suspect that the model portfolio lags behind the All-Share for two reasons: 1) The change from higher yield to higher quality stocks that I just mentioned and 2) an unsustainably high growth rate in the All-Share’s yield
As you can see, even over periods of five years it can be difficult to differentiate between luck and skill. Hopefully after another five years things will be a little clearer.
Low Risk: Less volatility and smaller declines than the market
My risk goal is to have the model portfolio fall by less than the All-Share in terms of maximum peak-to-trough declines over the last five years or more. Here’s the chart:
As you can see, the model portfolio had smaller maximum declines than the All-Share, and especially so during the All-Share’s recent 2015/2016 decline.
Or to put it another way, other than during a brief period during 2011/2012, the model portfolio has had no declines great than 4%, whereas the All-share suffered a significant double digit decline in 2015/2016.
With a but if luck this feature of the portfolio should hold true during the next major bear market, and I will be able to avoid the stress of seeing declines of 40% or more.
Although I invest in stocks where there is less corporate risk (i.e. the companies are more defensive and higher quality than average) and less valuation risk (the shares are trading on lower valuations than average), I think the main reason the portfolio has been less risky than the market is its policy of broad diversification.
Simply put, I make sure the portfolio is spread widely across different companies, countries, sectors and business cycles, where the goal is to have the short-term ups and downs from some holdings negating the short-term downs and ups of others. The net result, as Modern Portfolio Theory suggests, is better returns with less risk.
Here are some diversification-related charts:
Low Cost: Low turnover, low fees, low effort
The name “active investing” suggests that active investors are active, i.e. they are constantly buying or selling stocks; but that isn’t necessarily true and that isn’t necessarily the best way to go about it.
Many of the best active investors are actually quite inactive because inactivity has several benefits:
- It’s easier than being constantly active
- It’s less stressful
- It’s cheaper, because each trade can run up fees and taxes
- You can (perhaps) make better decisions as you have more time to examine each one
- It leads to longer holding periods which forces investors to focus on that are likely to grow over five or ten years, and these higher quality companies often produce higher returns with less risk
To stop me from becoming hyperactive I stick to a pre-defined plan of buying or selling one holding each month.
For me this is the right level of activity as it gives me something investment-related to do on a regular basis and stops me from becoming bored. It also helps to constantly improve the portfolio by regularly steering it towards better companies at lower prices.
During 2016 I sold six holdings that no longer scored well on the UKVI stock screen, and replaced them with top-rated stocks. The six sold stocks are listed below and you can click on the company name if you want to read the full post-sale review:
- Amlin PLC: Taken over in February for an 85% return over 3 years
- Hill & Smith PLC: Sold in March for an 83% return over 2 years 9 months
- Tesco PLC: Sold in May for a 38% loss over 2 years 11 months
- Reckitt Benckiser PLC: Sold in July for a 63% return over 2 years 5 months
- Chemring Group PLC: Sold in September for a 73% loss over 5 years 5 months
- Homeserve PLC: Sold in December for a 134% return over 3 years 3 months
The two big losers, Tesco and Chemring, were incredibly helpful investments, despite their large losses.
They were helpful because I learned several important lessons from each company, mostly about things to avoid such as excessive debts, excessive acquisitions and excessive capital expenses.
Despite the disappointment of losing money on those investments, I see it as cost which is occasionally necessary in order to improve the underlying investment strategy.
In fact, a sort of unstated fifth goal for the model portfolio is to provide feedback in order to drive forward that process of continual improvement, because continual improvement is the foundation upon which the other goals of high growth, high yield, low risk and low cost are built.
So much for 2016; what about 2017?
Rather than dwell on what may or may not happen in 2017 (let’s face it, nobody has a clue), I prefer to knuckle down and stick to the basics of: 1) Applying and improving my investment strategy, and 2) beating the All-Share over the next five years and more.
Are you sure that FTSE All Share is the representative benchmark for your portfolio?
I had a quick look at HSBC FTSE 250 and Dimensional UK Small Cap fund and they performed 98% and 113% in the last 5 years. I am not saying that any of those should be the benchmark but from what I know your portfolio carries a lot more stocks which are part of FTSE 250 and Smal cap indexes than in the FTSE All Shares.
As a result a benchmark that is skewed a bit more towards the FTSE 250 and Small cap would have made more sense.
In the end every portfolio could be decomposed based on the main factors that offer an equity premia: quality, small cap, value and momentum. If there is any outperformance then the investment manager could claim some skill.
Even that skill may be only linked with the period analysed. Peter Lynch when asked recently said that if he was an investment manager from 2010 onwards he would have underperformed heavily.
The same is with Bill Miller who beat S&P 500 in each of the 15 years up to 2006. That was the period when value investing worked and this guy was a real master at it, he was able to figure out what are good companies and what are value traps. No one will argue he did not have skill. However when his style went out of favour he underperformed 5 years of the next 6 and he was forced to retire.
This shows that even prolong past performance is not an indication for future investment return. It is just an indication that an investment style is working. I get worried when a style is working for a long time, that’s for me an indication for future underperformance.
Even Miller once said: “As for the so-called streak, that’s an accident of the calendar. If the year ended on different months it wouldn’t be there and at some point the mathematics will hit us. We’ve been lucky. Well, maybe it’s not 100% luck—maybe 95% luck.”
John Kingham says
Hi Eugen, that’s a really interesting point and one that has come up several times in the past.
I think benchmarking is a necessary because you have to compare yourself to something, but then the question of course is what?
I think a benchmark needs to be something that is investable and fairly similar (a deliberately ambiguous term) to the portfolio being benchmarked.
I think the FTSE All-Share is the simplest benchmark that will make the most sense to most readers of this website, although technically it is perhaps not the best benchmark.
The FTSE All-Share is made up of approximately 79% FTSE 100, 18% FTSE 250 and 4% FTSE Small Cap.
My model portfolio is made up of approximately 47% FTSE 100, 48% FTSE 250 and 5% FTSE Small Cap.
So you’re right to an extent; my portfolio has a higher weighting to the FTSE 250 than the All-Share does (although it isn’t significantly more exposed to small caps).
I think the best benchmark would be a 50/50 Split of the FTSE 100 and 250 as that would most closely match my portfolio with passive indices in a fairly simple and understandable way.
However, I have decided to stick with the FTSE All-Share as a benchmark because a) I still think it’s a very representative benchmark as my portfolio is fairly close to the All-Share’s make-up and b) the FTSE All-Share is investable as and a single ETF or fund while a 50/50 FTSE 100/250 benchmark isn’t.
Unfortunately these are inevitable problems with benchmarking and the same goes for the benchmarks chosen by any other investor, including Terry Smith and Neil Woodford.
No benchmark is perfect, so for me it has to simply pass the test of reasonableness, and I think the FTSE All-Share is a reasonable benchmark for my model portfolio.
As for lucky streaks, I tend to go by valuation more than anything else, and currently I’m quite comfortable with the aggregate valuation of my portfolio and the All-Share and FTSE 100 as well, although less so for the FTSE 250.
I remember you saying before that you were about to sell BAE systems, but its still in your holdings. Any reason why?
John Kingham says
Hi Andrew, yes there is a reason although not a very exciting one.
Basically I like to have at least half of the portfolio’s holdings in defensive sectors. Recently that’s been difficult to stick to due to a lack of attractively valued defensive sector stocks (if you look at the Business Cycle Diversity chart you’ll see that just over half of the portfolio is in cyclical sectors). So there is a balance to be struck between sticking to my rule on the number of defensive sector stocks and sticking to my rules about valuation (i.e. sticking to high-ranking stocks on my stock screen).
This month I should have sold BAE, but it’s a defensive sector stock and selling it would have left the portfolio even more skewed towards cyclical stocks. So, I chose to keep hold of it and sell the cyclical Tullett Prebon (or TP ICAP as it’s now know) instead. I’ll be posting the post-sale review of that next week.
However, BAE is still likely to be sold sooner rather than later, but exactly when I cannot say.