The discounted cashflow theory of everything
DCFs have a reputation for inaccuracy driven by high sensitivity to assumptions of company fundamentals and discount rates. To combat this, we do two things. Firstly, we limit the number of assumptions we use to two sets of variables, growth rate and fade rate, and use a deliberately limited menu of four discrete options for each. This eliminates the perennial temptation to over-fit a model and in doing so reduce its simplicity and comparability.
Secondly, we use a ‘reverse DCF’ which solves for the discount rate that reconciles current market cap and EV with our estimate of future levered and unlevered cashflows, respectively. This is analogous to a bond’s redemption yield.
We are careful to look at both levered and unlevered valuations. While the redemption yield on the equity piece is the best representation of outcomes available to equity investors like us, we are conscious leverage amplifies equity’s upside and downside.
We also use spot metrics as a cross check for calibrating the models. Our decision to concentrate on lifetime cashflow has been rewarding for our investors, with our flagship TB Evenlode Income fund realising a 4.5% excess annualised total return net of fees since inception in 2009, compared to its FTSE All-Share benchmark.
There are reverse DCF models available off the shelf for investors, and plenty of fund managers use them as part of a broader toolkit, often for specific companies at specific times. Where we are different is in two ways. Firstly, we built our own proprietary system which is driven by our EDDIE software platform, which allows us to scrutinise the model and audit its workings in detail. Secondly, we use this model for our entire investment universe. This comes from the thesis that while the model, like all models, will be wrong and approximate, it will tend to be wrong and approximate in a relatively similar way across our investment universe.
We thus get a clear sense of where there is value in our investment universe from redemption yields on offer. The alternative is to use spot metrics modified by our qualitative research on fundamentals to calculate the ‘right’ PER or FCFY for each universe company, but this, to our minds, is too unsystematic and complicated.
This leaves our expectations of future cashflow relative to current market cap and EV to stand alone as the gatekeeper to the portfolio. It is this process that demarks us in our sector and enables us to keep picking up reasonable businesses at the best prices available.
