Should I, shouldn’t I?

I ask this for the benefit of readers as much as for my own; the point here being that any serious investor must ask the same questions and run similar exercises to the ones I’m about to discuss with data from portfolio spreadsheets that they must maintain in order to keep on top of their investments.

When do random changes become a trend? Only when they are assembled with the benefit of hindsight. So it was hard to spot that in mid 2016 the income portfolio’s performance may be fading, especially as it was still adding value. Indeed, from mid 2016, when hindsight shows the fund reached a high-water mark relative to the FTSE 350 low-yield index, the income portfolio gained 21 per cent to reach what remains its all-time high in August 2017.

Still, as the chart shows, the income portfolio’s performance and that of the FTSE 350 Low Yield started diverging in mid 2016 and that divergence has since become significant. The table below puts that into figures. In the past five years the income portfolio has gained 5 per cent while the low-yield index has gained 33 per cent. The income portfolio has stood still in the past three years, while the low-yield index has climbed 18 per cent and, in the past year, the low-yield return has been almost 10 per cent to the income portfolio’s 4 per cent. This excess performance in the past five years feeds through to the cavernous gap between the two over the 10-year span.

Two caveats are important. First, the chart and small table are based solely on data for capital returns. They ignore the effect on total investment returns of dividends, of which the Bearbull portfolio gets a lot. As the table for the income portfolio at the end of October shows, over its 21-year life the fund has generated more from income received – £199,000 – than from capital value added – £197,000.

Most likely, if comparisons were based on total returns, the income portfolio’s performance would look much better. For instance, although I don’t have the data needed for a rigorous comparison, I can say that the portfolio’s average total return for the 19 years 2000 to 2018 is 10.2 per cent, almost twice the total return posted by the FTSE All-Share index (5.2 per cent).

The second caveat is that I am chiefly comparing the income portfolio with the low-yield segment of the FTSE 350 index, which focuses on growth stocks, whose capital returns over the long haul should be better than low-growth income stocks. True, that statement is contentious since income stocks are almost synonymous with value stocks and value – because it is serially underrated – can outperform growth for long periods. That said, in the past few years growth has clearly been in the ascendant. So why chiefly compare the Bearbull portfolio with growth rather than value? To make the comparison more demanding. To beat an easy target is to achieve little. As in anything we do, in portfolio management we need to push ourselves.

So, as I said earlier, what should I have done and what shouldn’t I have done? Taking bigger gambles would have been a fortunate move. That would have meant focusing on shares with dividend yields so high they implied disaster was around the corner.

Put it this way – back in late 2015, when the income portfolio’s performance was about to falter, the 10 highest yielding shares among the component stocks of today’s FTSE 350, excluding investment trusts, went on to produce an average gain of 61 per cent in the four years to today. True, among the 248 stocks for which there are four-year histories, nine of the 25 groups of 10 produced higher average returns, and the higher returns correlated with lower dividend yields. Even so, that 61 per cent gain, coupled with especially fat dividend returns, would have done nicely.

The Bearbull portfolio owned only one in that highest-yielding group of 10 – copper producer Antofagasta (ANTO), whose shares I had sold in mid 2015 when the company cut its dividend. There may have been others whose axed dividends would have made them unsuitable for an income portfolio. Against that, others among the 10 hardly looked distressed even four years ago, such as housebuilder Berkeley (BKG), insurer Hiscox (HSX) or gambling group GVC (GVC), then much smaller than the merger-driven sprawl it has become. Perhaps the investment moral – though stated tentatively – is not to be deterred by yields that look unsustainably high.

I might also ask – and contrary to what has just been discussed – if the Bearbull portfolio spreads its risk sufficiently widely? Now – as it was in mid-2015 – it is a 14-stock fund, which is not very many. Had the portfolio contained more holdings, its returns since 2015 might have regressed to the market’s average, which, in this period, would have been an improvement.

In general, however, I am looking for something better than market returns, for which more risk-taking – and a concentrated portfolio – might be required. Besides, the benefits of portfolio size on risk reduction can be exaggerated. Research has shown that ‘risk’ in a portfolio – ie, the standard deviation of returns from the average – halves as the number of holdings rises from one to 10. More important, there is no worthwhile risk reduction once a portfolio gets beyond 20 holdings. So a few more holdings in the Bearbull portfolio perhaps, but no more than that, especially as the time needed to research and maintain a bigger portfolio is its own constraint.

Talking of research, I might question the way that I dig out investment candidates. My off-the-peg approach focuses on number crunching from a company’s accounts. It uses past performance as the basis for guesstimating a range of per-share valuations – from optimistic to pessimistic – based on both accounting profits and cash flow. I back that up with more spreadsheet work to assess the trends in a company’s efficiency, its productivity and its financial resilience.

The merits of this approach is that it is an efficient way of scanning lots of candidates. Its shortcoming is that it pays insufficient attention to the future, which is where investment returns will come from. True, but the lion’s share of my time spent crawling over any company always comes down to relating the quantitative findings to the question, to what extent is the future likely to be as good as the past, better than or worse than? I don’t think that will change.

I could also ask if I am paying enough attention to stop-loss signals. Readers are fascinated by stop-loss techniques. I am less sure maybe because I am a believer in the efficient-market hypothesis. I reckon the chief merit of using stop-losses is actually to run profits – and lock in most of them – more than to limit losses.

Given that, I might wonder why the Bearbull portfolio still has a holding in touch-screens maker Zytronic (ZYT) since it was once running a 215 per cent profit on its investment. Part of the answer is because I did something different with Zytronic. I sold half the holding for a 123 per cent profit in late 2015, allowing me to run the remainder for a zero cost. For 18 months that looked good as Zytronic’s price surged through £6. Of course, the past two years have been lousy. Still, from one vantage point the Zytronic holding remains okay – dividends aside, and Zytronic’s payout now yields 11.8 per cent on cost – the capital gain is still 61 per cent of cost. Yet if I repeat the Zytronic part-disposal trick, I won’t treat a subsequent stop-loss signal as otiose.