Conservatism with a small ‘c’
William Lough highlights the importance of exercising conservatism within one’s investment process.
“[T]he growth-stock approach may supply as dependable a margin of safety as is found in the ordinary investment provided the calculation of the future is conservatively made, and provided it shows a satisfactory margin in relation to the price paid.” Benjamin Graham, ‘The Intelligent Investor’
In a previous note, I wrote about how we seek to find opportunities in the stock market where the true quality of a business is somehow hidden to the mainstream due to temporary factors. In this follow up, my goal is to highlight the importance of exercising conservatism within one’s investment process. It is this, above all else, which provides us with asymmetric payoffs. It protects us if we are wrong as well as tilting probabilities towards us of more favourable outcomes and therefore attractive returns.
It’s been said that ‘value investing’ is simply buying at a discount to what your calculation of intrinsic value is, based on the present value of forecast cash flows. This is true in principle, but also highly unsatisfactory from my perspective. There are two critical elements missing considering the application of conservatism.
First and foremost is the conservatism of any forecasts used. Discounted cash flow (DCF) valuation methodology is possibly at peak popularity today and is often slightly patronisingly pitched as the preferred methodology of the true connoisseur of fundamental investing. DCFs are a useful tool (which we use), but as with all models they are only as good as the inputs used … which can often be abused and reverse engineered to provide the answer the user wants! There are certainly plenty who deride DCFs for their ‘garbage in, garbage out’ nature. So, put simply, I believe that finding ‘value’ if your assumptions are based off high levels of growth being sustained – or even accelerating – far into the future and then using discount rates keyed off record low bond yields is stretching the definition, as the lack of conservatism creates multiple potential sources of disappointment. And secondly, if you buy at only a small discount to your estimate of intrinsic value then you put the odds further against you. Or, as Warren Buffett put it, “You don’t try to buy businesses worth $83 million for $80 million. You leave yourself an enormous margin. When you build a bridge, you insist it can carry 30,000 pounds, but you only drive 10,000 pound trucks across it.”
It’s instructive to use examples to show what good and bad practice looks like here.
Over the last 5 years, UK-listed companies Computacenter, Moneysupermarket.com and Hargreaves Lansdown all delivered a respectable 11-12% compound growth in EBIT (Earnings Before Interest and Tax). Presumably all showed similar share price performance, right? Wrong. Over the same period, Computacenter delivered a total return some 200%-plus (or 20-25% per annum) better than the other two.
The difference comes down to the starting point of expectations. Remembering that all 3 delivered the same attractive 11-12% EBIT 5-year CAGR (Compound Annual Growth Rate), consider the following:
- Moneysupermarket.com started the period on ~18x EV/EBIT (Enterprise Value/Earnings Before Interest and Tax), with a long-term EBIT forecast ~£130m or ~20% 5-year CAGR. In the previous 5-years EBIT CAGR was ~17%. They delivered ~£120m in 2019 (and are forecast ~£87m in 2020).
- Hargreaves Lansdown started the period on ~28x EV/EBIT, with a long-term EBIT forecast ~£360m or ~13% 5-year CAGR. In the previous 5-years EBIT CAGR was ~12%. They delivered ~£336m in 2019.
- Computacenter started the period on ~10x EV/EBIT, with a long-term EBIT forecast £100m or ~3% 5-year CAGR. In the previous 5-years EBIT CAGR was ~9%. They delivered ~£150m in 2019.
Hargreaves Lansdown and Moneysupermarket.com delivered attractive profit growth, but growth which disappointed elevated expectations within sell-side consensus forecasts – which implied a maintenance or pick up in profit growth over a 5-year period – or implied by the starting valuation multiple. Computacenter had a conservative expectation set and a relatively low hurdle for positive surprises provided by the valuation multiple.
Conservatism matters. Price matters.