Last week, we looked at James Montier's three-pronged approach to shorting stocks. Following on from that, it's also worth mentioning a useful accounting test that he developed in the context of shorting called the C-Score (C apparently stands for cheating or cooking the books). It's similar in nature to the Beneish M-Score, i.e. it's focused on identifying tell-tale signs / quantitative red flags which accompany bad accounting practice.
Even if you don't fancy shorting individual stocks (given the scope for unlimited loss!), the C-Score approach is still a useful tool as a red flag that's worth knowing about - as with the Altman Z-Score, we're considering adding it as part of the Stockopedia PRO Stock Report.
How does it work?
Montier's C-Score is made up of six red flags or warning signals. The idea is that, like the Piotroski F-Score for financial health, these elements are scored in a simple binary fashion, 1 for yes, 0 for no.
These scores are then summed across the elements to give a final C-score ranging from 0 (no evidence of earnings manipulation) to 6 (all the flags are present - yikes!). The areas tested are:
- Is there a growing divergence between net income and operating cash-flow? We've talked elsewhere about the fact that management have less flexibility to alter cash flow, whereas earnings can be stuffed for all sorts of quot;funniesquot;, so this is something to watch for.
- Are Days Sales Outstanding (DSO) increasing? If so (i.e. accounts receivable are growing faster than sales), this may be a sign of channel stuffing.
- Are days sales of inventory (DSI) increasing? If so, this may suggest slowing sales, not a good sign.
- Are other current assets increasing vs revenues? As some CFOs know that DSO and/or DSI are usually closely watched, they may use this catch-all line item to help hide things they dont want investors to focus upon.
- Are there declines in depreciation relative to gross property plant and equipment? This guards against firms altering their estimate of useful asset life to beat earnings targets.
- Is total asset growth high? Some firms are serial acquirers and use their acquisitions to distort their earnings. While this may be justified in some circumstance, generally it has been shown that high asset growth firms underperform. It's not clear what Montier defines as quot;highquot; but we're assuming that firms with above average asset growth would qualify as 1.
Of course, any of these elements on their own may be perfectly innocent, but it's probably worth knowing the explanation for it before you invest. The idea is that, the more flags that are present, the more likely it is that something may be going on below the surface of the accounts. Having said that, Montier's original list from 2008 contains some fairly respectable names like Amazon - so beware of false positives, as with any statistical measure.
Does it work?
Montier found that, in the US, stocks with high C-scores underperformed the market by around 8% p.a., generating an absolute return of just 1.8%. In Europe, high C-score stocks underperformed the market by around 5% p.a., although, interestingly, they still generated absolute returns of around 8% p.a...
As a shorting tool then, Montier suggests using the C-Score in combination with some measure of over-valuation. This was on the basis that high-flying and generally more expensive stocks that are tempted to alter their earnings in order to maintain their high growth status. He used a threshold price to sales ratio of 2 and found that this drove the absolute return down to -4% in both the US and Europe!
Montier's excellent book, quot;Value Investing: Tools and Techniques for Intelligent Investmentquot; is available on Amazon (Chapter 25 discusses this strategy - it is also available online).
If you’d like to run quantitative long and short screens like this across the UK market, sign up now for beta access to Stockopedia PRO, our UK stock screener.
- Miguel Barbosa Interviews James Montier
- TMF: Where you'll find the next killer short
- Montier Presentation: White Swans, Revulsion amp; Value