Buying assets for less than they are worth is fundamentally at the heart of most sensible investment strategies. However, approaches to identifying value can range widely, from direct attempts to value businesses to approaches that exploit the systematic behavioural errors that other investors make (resulting in ‘factors’).

In my view detailed valuation exercises are generally too arbitrary and difficult and so are a bit of a waste of time compared to basing a strategy on statistical factors (e.g. quality, value, momentum). However, some simple valuation modelling can have incremental value if you understand its limitations and get the rest of your investing framework right.

How should you think about valuation

A point that is in my view underemphasised is that the proper valuation of a business is necessarily about forecasting the future – valuation depends on the value of future profits turned into cash and dividends.

Accounting for growth correctly is therefore an essential part of valuation. The quality and likely growth of a business are as important as the current profits and price (or the price earnings ‘PE’ ratio) in determining valuation.

This reminds me of a quotation that used to be on the menu of our local Indian restaurant:

The bitter taste of poor quality lingers long after the forgotten cheap price.

Amusingly, the restaurant in question didn’t really make clear that its dishes were of high quality, while many of the prices on the menu were more obviously cheap, making it unintentionally read like a warning not to eat there.

The importance of quality and compound growth in valuation is generally well recognised, for example by Warren Buffett, who is often quoted as preferring a great business at a good price rather than a good business at a cheap price.

The fact that profit growth can compound over time is very important. Small changes in the growth rate can imply very large differences in valuation when you look further into the future, compounding growth over a number of years. Of course, looking further into the future introduces greater uncertainty, and the impact act of a bad year or two on the compound average long term growth rate can be severe. This is why defensive companies with strong moats and stable consistent growth are so valuable.

Because true valuation is about the future it is also inherently somewhat arbitrary. It needs to account for one’s attitude towards risk and preferences over time (or discount rate). Because of this, the valuation of a share relative to another is more meaningful than valuation in any absolute sense. Therefore the key attribute of a valuation metric in my view is to rank valuations relative to one another correctly, rather than to tell you whether the market as a whole is undervalued (which is really a matter of perspective – I think there is a pretty strong case for saying that equities are very rarely collectively overvalued if you take a reasonably long term perspective).

Ways to measure value

The most direct way of incorporating compound growth into valuation is to use discounted cash flow (DCF) modelling. This looks at the future streams of cash coming into the business and discounts them according to how far in the future they occur (to reflect that a £ now is worth more than a £ in one year). The disadvantage of DCF modelling is that it is subject to a lot of uncertainty and the results are are very sensitive to the assumptions that need to be put in to it.

In this context, I think there is little merit in trying to construct a complex DCF model. This results in a lack of transparency that is not helpful. Simple ratios that can give some idea of whether something is undervalued but don’t give a false sense of accuracy are more practically useful. These can then be used in a overall assessment in conjunction with other factors, rather than trying to bake all the assumptions into the model.

In thinking about the most useful valuation ratio or model, there is a trade off between the accuracy of the ratio as an indicator of the true valuation (how much is ‘baked in’ to the model) and its simplicity and transparency (as other factors can be assessed outside the model).

I think the best option is something in the middle, that is still simple and easy to interpret, but that does reflect what the impact of likely compound growth on valuation would be. I think Benjamin Graham’s rule of thumb does this pretty well. To understand why, I think it’s easier to first go through the strengths and weaknesses of the more common and basic valuation ratios: the PE ratio and PEG ratio.

The PE ratio

The PE ratio of a business is simply its share price divided by its current profits per share.

While few trust DCF modelling, many appear happy to put a lot of weight on PE ratios in making decisions. Don’t get me wrong – if used correctly I think the PE ratio is a very sensible starting point that has many attractive qualities: it is simple, easy to use and understand what it means.

As I’ve hopefully made clear above, its main drawback as a valuation metric is that it doesn’t account for growth. This matters a lot if you want to invest in high quality businesses that are growing. It is of course possible, when deciding what share to buy, to mentally adjust for growth prospects and quality separately to the valuation metric. It is not difficult to understand that a business with better prospects should be worth a higher PE ratio. However, it is not obvious how large this mental adjustment should be.

In my view the main advantage of the PE ratio is also its main weakness. Because it is so easy to understand and use, it seems to have a disproportionately large impact on decision-making. It is relatively easy to compare the PE ratio across businesses but relatively hard to compare quality or growth prospects, while really the differences in the latter tend to be a lot more important.

If this all seems a bit obvious remember the impact is subconscious too. If your decision making process involves an element of gut-feel and assessing lots of points in the round (as most do) then it is easy to end up over relying on the PE ratio whether you realise or not. For this reason I try to avoid looking at it as much as possible.

PEG ratio

The PEG ratio incorporates growth into valuation in a straightforward way by dividing the PE ratio by the earnings growth rate. The idea is then that a lower PEG ratio reflects a valuation.  People often use a mental benchmark of 1 as reflecting fair value – so if a share’s earnings growth rate is greater than its PE ratio then it is ‘cheap’.

You need to be careful about the growth rate used to calculate the PEG. Most often the one year earnings forecast is used to calculate it. However, using the current growth rate over one year seems a bit foolish if the idea is to approximate the true valuation. I think it is a much better idea to use a more long term growth rate and this is rarely reflected by a single year forecast. For example, companies that are experiencing explosive growth will have tend to have this growth moderate. Estimating a more long term growth rate is difficult but there is no getting around it – in fact it is the main underlying driver of the true value of a business.

I don’t ever use the PEG. My main criticism is not that it uses the wrong growth rate, as this is easily remedied. It’s more that, while it uses the pertinent bits of information, the formula seems arbitrary and more importantly dividing by the growth rate doesn’t seem right and leads to some dubious results.

In particular, it suggests a linear relationship between the growth rate and the PE ratio a share should trade at which seems obviously wrong. For example, it suggests that a share with twice the growth rate should always trade at twice the PE ratio. While this is true for one particular growth rate (I think a share growing at 20% for 10 years is probably worth about twice one growing at 10% for 10 years), for other growth rates it is clearly wrong. A share growing at 4% for 10 years is clearly worth less than double one growing at 2% for 10 years.

Graham rule of thumb

The Graham rule of thumb formula is a similar idea to the PEG but in my view is a better approximation. It is given by:

V = EPS * [8.5+2*G]

Where V is the intrinsic value, EPS is earnings per share and G is the growth rate as measured over the medium term e.g. 10 years.

If you divide both sides by the EPS, you can translate it into a PE valuation. The formula suggests a share should trade at a PE ratio of 8.5 +2*G. So a share with no growth should trade at a PE ratio of 8.5, one growing at 5% a year for the next 5-7 years should be valued at a PE ratio of 18.5 and one growing at 10% should be valued at a PE ratio of 28.5.

I haven’t found how this formula was originally derived, but it can be derived from a simple DCF model making assumptions for the discount rate. Here is a link to an article that discusses it alongside other similar metrics and another that compares it to DCF. What I like about it is that it is simple enough to calculate in your head, while also giving a reasonably accurate first order approximation for valuation (as a DCF model would), combining the PE ratio with the growth rate. It’s not perfect but the results seem sensible and I haven’t found a better alternative.

Putting it into practice

In a sense it doesn’t matter that much about having too precise a formula, as the results are to a very large extent driven by your forecast for the 5-7 growth rate.

Forecasting anything is hard and a growth rate is no exception. I try to be cautious in my assumptions, be very careful with using the forecasts of analysts and most importantly only try to value businesses which I think are high quality i.e. businesses which for which I think the extent and certainty of growth is high. For the majority of businesses whose quality I’m less sure of, there is simply too much uncertainty about future growth to have much confidence in my valuation of them.

Another factor that makes estimating the 5-7 year growth rate hard is that it may vary over those years. For example, consider the impact of one bad year on the valuation of a business that would otherwise grow at 10% a year. Say instead you expect growth to fall by 10% one of the years. How much should this affect its compound average growth rate for the 5 years? The answer is maybe by more than you would expect – it would have an average growth rate of only 5.6%. Of course the easiest way to avoid this is complication is to avoid the types of business where this is more likely. There is always risk though and this is a good reason to employ cautious estimates.

Another typical situation I often face is trying to value a business whose current growth rate is exceptionally high but likely to moderate over time. I find that it is worth working through example growth trajectories and seeing what average this leads to, as the effect of compounding can sometimes make this hard to guess in your head.

To illustrate this I’ll work through a couple of real examples from my portfolio: Fevertree and On the Beach.

Fevertree

Fevertree has just announced interim results with earnings 113% ahead of last year. Extrapolating this to the full year would put it on a current forecast PE ratio of 45. The current average analyst earnings forecast puts it on a PE ratio of 63 but this doesn’t seem to reflect how well they have done at the interim stage so looks some way off the mark to me. I’m going to assume a current forecast PE ratio of 50.

Given this year’s growth I’ve looked at the following growth trajectories:

• Years 1-5: 50%, 35%, 25%, 15%, 15%
• Years 1-5: 35%, 25%, 15%, 15%, 15%
• Years 1-5: 25%, 20%, 15%, 10%, 10%

They are progressively more cautious and would result in average compound growth rates of 27%, 21% and 16%. These in turn would result in ‘fair value’ PEs according the Graham formula of 62.5, 50.5 and 40.5. As the current forecast PE ratio is about 50 this seems to sit slap bang in the middle of my estimates and suggests Fevertree is about fairly valued. This will change over time depending on how well it performs against the possible trajectories mapped out.

The other way to use the Graham formula is to reverse engineer the implied average growth rate from the current actual valuation. In this case a PE ratio of 50 implies a growth rate of 20.75%.

On the Beach

On the Beach has a forecast PE ratio of about 21, based on analyst forecasts for growth this year of 60%, which in this case seem reasonable to me. I’ve tested the following growth trajectories:

• Years 1-5: 40%, 20%, 20%, 15%, 10%
• Years 1-5: 30%, 20%, 15%, 10%, 10%
• Years 1-5: 25%, 15%, 0%, 10%, 10%

Again they range from most optimistic to most cautious, the third one allowing for one year of zero growth. The compound average growth rates are 21%, 17% and 12%, implying fair value PE ratios of 50.5, 42.5 and 32.5. All of these suggest On the Beach is substantially undervalued. The implied average annual growth rate is only 6%.

It is important to recognise that these are really just thought experiments based on broad assumptions. What hopefully comes out from them is a better understanding of the scale of the uncertainty surrounding the valuation and a broad sense of how undervalued or not a stock may be.

Loose ends

Valuation in reality isn’t just about profits and growth. There are some other important relevant factors such as the net debt and ability of the business to turn profits into cash. Just a couple of final comments on these to wrap things up.

How do you value the net cash or debt on the balance sheet? Conceptually this is straightforward – £1 on the balance sheet is worth £1! So really you should net this off from the valuation before you start. The normal way to do this is to use the enterprise value EV, which is calculated as the market cap plus the net debt (it is the implied value the market placed on the underlying business stripping out debt). I tend not to bother doing this out of laziness, but then again pretty much all of the business I invest in generate a lot of cash, tend to be able to grow with little debt and consequently have no net debt or occasionally a substantial amount of net cash.

Similarly, the valuation metrics above don’t take account of the ability to turn profits into cash, or to put it another way how much investment is required to drive growth. This is critically important but my approach is to look at this separately to valuation when assessing quality so as to keep things simple.