I’m working on changes in client portfolios (and mine as well, they mirror my portfolio; I eat my own cooking!), and I spend time looking at longer-term returns on a book value basis.? I agree with Buffett; growing book value per share grows market value per share over time.
I write this because I often see companies that are cheap on an earnings basis, but have been so for a while, but have not grown book value per share (after reinvesting dividends) by much.? This comes from non-operating losses and badly-timed buybacks (or, persistent buybacks that don’t take account of current market price).
You can have a business that throws off great free cash flow, and waste it by buying back stock when the market is willing to pay too much, and the company intensifies the mistake.? Better to build up cash, and wait for a better day to buy, when shares are cheap and worth buying.
P/E is a flawed measure because few ask what is done with the E.? Is it used to good ends?? We need to look further and analyze how excess cash gets used for growth.? Stock buybacks are a minimum, but maybe waiting to do stock buybacks at lower prices is better.? Maybe buying out small private companies that can round out a product portfolio are better still.
Regardless, the goal should be to grow book value, adding back dividends.? That is a path toward sustainable growth in stock values.
I rank by something that sounds similar. I look at the change in total equity minus intangibles and goodwill over a period and divide it by the average retention rate for that period, the idea being to check whether reinvestments have been effective over the period.
I don’t find it so helpful alone but when paired with a value metric like earnings yield, the combination has worked better than either alone.
I like your blog. Keep up the good work!
Al
Thanks very much – this is interesting. The reason I often haven’t looked at book value is because of skepticism of how many companies value the assets on their books (over-valued or silly depreciation rates). However perhaps as long as the methodology remains consistent, looking at changes in BV/S still remains valid… will start taking a closer look at this. Thanks, Stefan