Thursday, 3 March 2011

Lies, Damned Lies and...Financial Ratios

Following on from a previous excursion into Return on Equity, and the very good debate which ensued on this blog and on Stockopedia, I have been pondering further on how to identify and recognise what constitutes a great company.

Why Great?

It sounds like a redundant question, but why the interest in great companies?

To paraphrase the likes of Buffett and Co, they would rather buy a great business at a fair price, rather than a fair business at a great price. This makes perfect sense as the stronger a business is (and is likely to remain so in the future), the stronger the likelihood that future profitability will increase. Get the quality business at a discount and then you can make a killing a la Greenblatt.

I may have come to it a bit late, but there seems to have been a lot of excellent analysis and discussion around the subject recently, particularly with UK Value Investor's screening based on ROE and other attributes, and Richard Beddard's Herculean efforts in attempting to apply Greenblatt's Magic Formula to the UK with ROA being a key component.

I have been taking a closer look at other ratios that are commonly used alongside ROE.

Useful Tools for the Tool-kit?

1 - Return on Assets (ROA)

ROA = Adjusted Net Income / Total Assets

Adjusted Net Income = PAT + Interest (and the tax effect) - to show (kind of) operating profits without the cost of the capital base (ie interest)

Total Assets = Net Assets (opening or average)

Observations. Trying to adjust for interest and the tax effects is fiddly (Sharelockholmes in its ROTA calculation gets around this by using EBIT (I think)). Net Asset Value will include book values of intangibles and goodwill, and fixed assets, all of which may be very detached from market value.

2 - Return on Capital (ROC)

ROC = EBIT / (Net Working Capital + Tangible Assets )

NWC = net current assets/liabilities - ie stock + debtors + cash - current liabilities. Cash should be operating cash

TA = opening or average book value of Tangible Assets

Observations. This is very similar to Greenblatt's Magic Formula measure, although adjusting for surplus cash remains problematic. At least intangibles and goodwill are excluded from Assets, but there is still the issue of tangible assets at book cost. Maybe a high ROC is a sign that the assets could be undervalued?

3 - Return on Capital Employed (ROCE)

ROCE = EBIT /  (Total Assets - Current Liabilities)

Very similar to ROC but includes long-term debt.

So What?

They all sound fine and dandy and they have particular strengths and weaknesses, but it is too easy to get hung up about the finer details of these ratios in isolation. Every man and his dog seem to tweak each definition, and each sector has different drivers which will influence the outputs when compared to other sectors. Formulas are useful, but should not be the 'be all and end all' and do not remove the need to understand the movements in a company's finances. At the end of the day, we're all try to find good quality companies at an appropriate price, and I'm going to use them as an early stage screening tool.

I am in the process of re-writing/tweaking my Rules, but am considering using the following measures for screening purposes:

1 - ROE > 15% average over 10 years 
2 - ROC > 25% (as suggested by Greenblatt - ROA equivalent)
3 - ROCE > 15%
4 - PER < 14
5 - Yield > 2.5%
6 - Gearing < 50%
7 - Mkt cap > £20m

A first pass has thrown up 19 interesting companies, including the likes of Astrazeneca and Unilever. In an ideal world, I would want to back-test these parameters so see what the results are like, but for the time being I will judiciously select a few targets and take them to the next level of analysis. Watch this space...


  1. You're right Yorkiem there is no perfect profitability measure. Doesn't seem to stop us trying to find it though! I hadn't thought of combining them like you have.

  2. Thanks Richard. Do you offer an outsourced back-testing service?!

  3. Another Smart post from you Admin :)