Tuesday, 24 July 2012

The shocking truth about growth investors ... is that they are right.

The shocking truth about growth investors

Growth investing and growth investors are dirty words to many value investors.  A focus on growth is often called the Greater Fool Theory, and study after study shows that growth investing performs badly when compared to value investing.
As a value investor, it’s easy to mock growth investors with reams of data and an air of self-satisfied superiority.  There is a slight problem though.  The shocking truth about growth investors is that they’re right.  Growth investing is a fantastic way to make money in the stock market, as long as you do it right.
Warren Buffett is a growth investor
Buffett is usually considered a value investor, and that’s because he is one.  But he’s also a growth investor, and with the help of Charlie Munger they pioneered a hybrid approach where they combined the best of both worlds – long-term growth companies bought at value investment prices.
It was Buffett’s focus on outstanding businesses which could grow both quickly and consistently that really took him to the top of the world’s richest people list.
The FTSE 100 as a long-term growth investment
As a UK investor my focus is always on beating the FTSE 100 in the long-run.  The FTSE 100 beats some 80% or so of private and professional investors alike, and so if I can beat the FTSE then I know I’m doing much better than the pros, which is always nice.
I also know that my time and efforts are not wasted, because any investor can invest in the FTSE at almost no cost in terms of either time or money.  If you are not beating the FTSE 100 then you are effectively wasting your time.
As a growth investment, the FTSE 100 typically grows both earnings and dividends faster than inflation, and it does so relatively consistently over the years.  That growth ultimately drives the index level higher, regardless of how pessimistic the market may be.
In order to beat the market, we need to turn our portfolios into supercharged versions of the index, with superior growth, superior yields and superior valuations.
We all want growth, but growth of what?
For me, the most important numbers that need to grow are revenues, earnings and dividends.
At the end of the day, it’s the earnings and dividends which set the range within which a share price will fall (exactly where it falls within that range is up to the market), and both of those ultimately derive from revenues.
Long-term growth is all that matters
Short-term growth, positive or negative, is mostly noise and is unlikely to provide any useful information to investors.  If you find yourself trying to make money out of the day-to-day news then you might have inadvertently become a trader rather than an investor.
When I’m talking about growth, I mean long-term growth over as long a period as you can sensibly get data for.  For me this means looking at 10 year data for every single company that I’m interested in, and if it doesn’t have 10 years of public data available, then I won’t touch it.
This means that Facebook was out of the question, no matter how attractive it may or may not have been.
Where to get your data
Getting data that goes back 10 years can be tricky, but it is available through services likeSharelockholmesMorningstar PremiumShareScope and Stockopedia.
You can also get the annual results yourself and copy the information into a document or spreadsheet and use that.  I like to get annual report data from investegate.co.uk because it has a nice, lightweight and fast interface and I’m used to using it.
One slightly odd feature is that nobody seems to provide revenue per share.  We always get earnings and dividends per share, but not revenue per share.
If you want revenue per share, as I do, then you can either ask a data provider to provide it or get hold of the number of shares outstanding figure for each company you’re interested in.  Just search the annual reports for ‘shares’ and it should appear somewhere.
Of course you can always get your initial data from my newsletter, the Defensive Value Report, which gives high level data such as PE10, G10 and yield (all of which I’ll cover in upcoming posts) for all FTSE 350 companies.
Okay, so let’s get into the details…
How to measure revenue growth
The simplest way to look at 10 year growth is to just compare the revenue per share figure from the latest annual report with the one from 10 years ago.
I’ve tried various combinations to find the most accurate and robust measure of long-term growth.  I’ve tried looking at the average of the growth from each individual year, or combining 10, 5 and 3 year growth rates, but after much experimentation it seems that a simple 10 year growth figure is as good as anything else for highlighting long-term growth companies.
One caveat with revenue is that some companies don’t have revenue numbers.  Depending on where you get your data you may not see revenue for banks, insurance companies and various other types of businesses.  In these cases you’ll just have to look through the annual reports to work out what the equivalent of revenue is.  For example, with insurance companies I use net written premium.
How to measure earnings growth
For earnings I prefer to look at adjusted earnings.  The reason for this is that I’m looking to measure growth over time, so I need a reasonably smooth and less volatile number to measure, and adjusted earnings tend to be less volatile than basic earnings.
With basic earnings, even in very stable businesses you can have big changes in a single year, or even losses which will mess up any long-term growth calculation.  For example, if the loss doesn’t impact the company’s long-term earnings power.
Earnings power is a term that I like because it conveys the idea of a company’s ability to earn money, not just the actual amount that it earns in any one year.
If you look at BP for example, then the last 10 years basic earnings look like this:
BP Table
So in 2010 there was a big loss in basic earnings.  If we were to take the 10 year growth figure in 2010 we’d have a negative 10 year growth number for that period, which would be hugely misleading.
This is less of a problem for revenues because that’s a more stable number and is never negative.  With dividends the problem does exist, but to a lesser extent because dividends are typically more stable than earnings.
By looking at adjusted earnings instead of basic earnings we can get a clearer picture of what the company is actually doing, and how the earnings power may be changing through the years.
But we can go a step further.  Ben Graham came up with a scheme for reducing the volatility of earnings even more, giving perhaps an even better picture of how the company’s earnings power is changing.
Graham simply took the latest 3 year average of earnings and compared that with the 3 year average from 10 years ago.
To get the 3 year average from 10 years ago you’d need the data going back 13 years (as the earlier average would be from years 13, 12 and 11).  If you only have access to 10 year data then you can just use the same system but just using the earliest 3 years that you have, which actually gives the 7 year growth rate between the two 3 year averages.
In the BP example above, we’d compare the average of 15.64, 22.88 and 36.48 (which is 25) to the average of 45.49, 77.48 and 79.04 (which is 67.34).
The growth over that period is 169%, according to my spreadsheet.
And talking of spreadsheets, if you want to know the annualised growth rate over that period you can just use the rate function in excel, which would look like this:
Where 7 is the number of years, -100 is the ‘present value’, and 269 is the future value (i.e. 100 plus the 169% increase).
The answer is that the 3 year earnings power of PB grew by an annualised rate of 15.2% per year in that 7 year period.
How to measure dividend growth
Like revenues, dividends are generally more stable than earnings, especially with the kind of large, market leading, relatively defensive companies that I’m interested in.
For that reason I generally just use the 10 year growth rate in the same was that I do for revenue.
However, I’m always experimenting with different ways of measuring past performance.  I want the most accurate and robust methods for finding companies that can grow quickly and consistently over many years.
That may mean that at some point I might change my dividend growth measure, and if it does it’s likely to change to the same approach that Ben Graham suggested for earnings.
Putting it all together
I call my growth metric G10, because otherwise it’s a massive mouthful to say that it’s the average of the 10 year growth of revenues, adjusted earnings and dividends, where the adjusted earnings growth is calculated as the growth between the latest 3 year average and the 3 year average from 7 years ago.
Just because this is a relatively complicated measure of growth, it doesn’t mean that it has magic powers.  It’s just as likely to throw up anomalies and rubbish companies as any other numbers based approach.
However, it’s a sensible first step towards finding companies that can grow earnings and dividends faster than the market, consistently and over long periods of time.
What it doesn’t really address is consistency.  So for consistency I have a separate metric which I’ll cover in my next post.

No comments: