Lifting the Hood on Performance

In our June 2017 piece, What to Expect from a Good Portfolio, we demonstrated how the ultimate driver of long-term performance for a static portfolio is Sustainable Return – i.e. the fundamental growth of the underlying businesses, measured in terms of revenue, earnings, etc., including capital returned to shareholders. We also demonstrated how even the best portfolio should be expected to underperform the market at times along the way to delivering superior long-term return (because randomness in valuations is what drives stock performance in the short-term).

While our portfolios in aggregate have delivered satisfactory mid-teen USD returns in 2017, they have lagged the benchmark index for the year (vs outperformance in 2016 / in-line performance in 2015), mostly because of our cash holdings1. As we demonstrated in June, this is by no means inconsistent with our expectations for short-term returns from a good portfolio, but it would nevertheless be reassuring to see some hard evidence that we’re on the right track to achieving our goals2. So how do we do this? We start with the end in mind:

Our primary goal is to maximize total return over the long-term (5 to 10 year periods)3. Total Return = Sustainable Return + Revaluation. In order to maximize return, we need to buy good businesses (high sustainable return) cheaply (implying positive future revaluation) and give them 5 to 10 years for the sustainable return to compound and for the valuation to normalize and perhaps overshoot. We have little doubt that if we do this consistently across a globally diversified portfolio, we will achieve our investment goals.

Given that we’re nearly at the 3 year mark (roughly halfway) we can’t read too much into our total returns, good or bad. Instead we need to look for evidence of a) superior sustainable return, and b) attractive valuations. The numbers are as follows:


Over the last 3 years, we estimate that our portfolios (excluding cash) have delivered total USD sustainable return of 18.8% vs the benchmark index at 6.7%. Our estimates for future sustainable return also comfortably exceed the index, as we would expect them to.

In terms of revaluation, our portfolios in aggregate are trading just below our estimates of fair value4, while the global stock market as a whole is in expensive territory. If we compare the valuation of our current portfolio to the benchmark index for the last 20 years, we find ourselves at the 15th percentile of relative valuations – i.e. relative to the MSCI World Index, our portfolio valuations are very attractive.

Even though the benchmark index has performed well over the last 3 years in terms of total return, its sustainable return has only been 6.7% in total. The rest has come from revaluation from expensive territory to even more expensive territory, which isn’t sustainable. Weak sustainable return combined with overvaluation is unlikely to deliver satisfactory returns over 5 to 10 years.

In contrast, Bellwood Capital’s portfolios demonstrate hard evidence of superior sustainable return and attractive valuations. Excluding cash, our portfolios in aggregate have performed roughly in-line with the benchmark index since inception, but have done so on the basis of more sustainable fundamentals, which remain intact. We are comfortable that our portfolios are well on track.

Another important consideration is that our portfolios are dynamic, not static. We don’t buy a portfolio of stocks once-off and leave it, which is basically what an index tracker does. Instead we aim to buy stocks cheaply, and sell them if they become too expensive, replacing them with cheaper stocks again5. Over time, a static portfolio, an index or an individual stock can’t be expected to outperform its sustainable return, because any revaluation should cancel itself out over time. A dynamic portfolio can outperform it’s sustainable return over time if it consistently takes advantage of revaluation by buying cheaply and selling expensively.



1 Read more about our reasons for holding cash here.

2 For the same reason that short-term underperformance is not inconsistent with a good portfolio, short-term outperformance is also not inconsistent with a bad portfolio. Short-term returns are random. Rather than assessing performance on this basis, we should be looking for evidence that long-term fundamentals are in place regardless of whether we’ve over- or underperformed short-term, something far too few investors do.

3 The reason that our goal is to maximize returns over 5 to 10 year periods rather than 1 to 3 year periods is because revaluation is the dominant driver of short-term returns and it is highly unpredictable. Revaluation is driven by human emotion, and the only reasonably predictable thing about human emotion is that it is likely to swing erratically between the varying extremes of optimism and pessimism over time. The more time we allow ourselves, the more likely we are to see pessimism develop into optimism, and vice versa. Further, as time passes sustainable return starts to compound and become the dominant driver of return. History suggests that in most cases 5 to 10 years is required for both of these factors to play out.

4 Over time we would expect our portfolios to tend towards fair value, with a slight undervalued bias. A new portfolio would be strongly undervalued on day one since we would be buying businesses cheaply. As time passes most of these businesses should tend towards fair value, and some would overshoot, resulting in a fairly valued portfolio on average. The stocks that overshoot too much would be replaced with new undervalued stocks, resulting in a slight undervalued bias overall. That our portfolios have moved from cheap to almost fair in 3 years is in line with these expectations.

5 Buying cheap and selling expensive is usually a slow process that plays out over years, and shouldn’t be confused with short-term trading for quick gains.


Past performance is not an indication of future performance. Equity portfolio values are subject to volatility and can increase as well as decrease.