Warren Buffett uses his retained earning test, where he looks for every $1 of retained earnings to become at least $1 market value. This is a novel idea, basically saying that a business should be wary of the returns on the money they’re not paying out in dividends. However, market valuations are not based efficiently or rationally in many cases (my Aeropostale investment still hasn’t paid off because cotton prices continue rising & Wall Street isn’t focusing on the underlying business). So, I’m writing this assuming market valuations are wrong (as they do sometimes end up far too high or low).
Think about this- a P/E ratio of 8 would require a 12% return on retained earnings (whether in machinery, inventory, marketable securities, etc.). $1 of investment turns into $0.12 in increased earnings, causing the market value of the company to go up 8*$0.12, or $1.00. If a firm has a P/E ratio of 15, it needs a 6.7% return on retained earnings to attain the same dollar of market value (6.7% * 15 = $1). With different P/E ratios assigned to different companies (which is far from scientific), it seems this crude approximation is not the best way for the advanced investor to think about reinvested earnings. Of course, Buffett was drawing this distinction so that it would get people to start thinking about returns on capital, rather than just growth for growth’s sake.
So, I figure I’ll put this in my context for the offshore drilling industry. At some drilling companies, where they are earning 15-20% return on invested capital but usually are trading at lower P/E ratios, market values go up anywhere from $0.75 to $1.00 in relation to incremental earnings. Although many of these drillers don’t qualify under his method, it still makes sense to think about absolute returns on capital. These drillers will be a great place to have some of your portfolio in the future (Noble is still the best in my mind). Investors’ idea of specific multiples to be attached to companies should not affect the ultimate expected return on retained earnings or future income that will be derived from the business.
Although this whole concept is a novel idea, why not just require a rate of return exceeding that of what an investor can reasonably expect themselves to accomplish in their own portfolio? If one has achieved 15% annual returns in his portfolio, it would make sense to require the return on retained earnings to equal at least 12% or so (remember the taxes on dividends & needing to find new places to put your money). That way, you end up thinking about certain businesses you don’t want to hold long-term because low returns on capital won’t make you wealthy over time. You can buy & sell the bad businesses if they get to be severely undervalued, but buying & holding mediocre businesses for a long time isn’t the formula for success.
So, when you ask me why I made my cutoff is 12% or so from an ROA standpoint on paying up for growth, this is why. I think I can reasonably achieve about 15% annualized if I work very hard, so 12% seems good to me (although 11% is still fine- the trick is staying away from the consistent 7% or 8% ROA for long-term investments).