Introducing the Boss Score part 4 – How it works in practice (KO, MUV2, Jumbo)
After my previous posts about the “Boss Score”, I think I had lost some of the readers. So I try to show with some hopefully interesting examples how the score works in more detail.
Low Price /Book ? – example Coca Cola
One reader commented that the strategy is designed to look for low P/B stocks. This is not the case. Lets have a look at CoCa Cola, maybe one of the most famous Warren Buffet / Moat style investments.
Coca Cola trades at a P/B of 5.2 and a PE of ~20, so how does this work in the screener ? In the current “simple” version without share repurchases, the database shows the following 10 year history:
|Book per Sh||Div||TROE %|
This results in a 10 year average TROE of 29% and a standard deviation of 9.8%. If we plug this into the CAPM formula, Coca Cola gets an adjusted (fundamental) CoC of 5.4%. Assuming now a constant 29% ROE based on a book value results in an assumed 4.21 USD constant earning per share going forward.
Divided by the 5.4% CoC, the “fair Value” of Coca Cola turns out to be ~74.80 USD, -0.8 below yesterdays close. The corresponding “boss Score” is -0.01, indicating a “fair valuation“.
So one clearly sees that the model is reacting very positively on high ROEs, solid balance sheets and stable owner’s earnings. However by design, any future growth is not included. So CoCa Cola might be even a great investment if one is sure that they will grow going forward.
Low P/B Financials – Example Munich Re
Munich Re has become a favourite for many conservative investors. It trades below book, has a low P/E and pays a good dividend, so what could go wrong ?
|Book||Div.||“Owner Earnings”||Stated earnings|
If we look at the comparison of “owner’s earnings” and stated EPS, we can see the problem clearly: EPS only reflect part of value creation or destruction. Munich Re is booking a lot of negative stuff directly into equity, thanks to the benefit of modern IFRS accounting.
Over the last 10 years, Munich Re reported a total of ~93 EUR in EPS, but summing up equity and dividends, the owner only received 62 EUR. This in turn leads to a relatively miserable TROE of 5,1%. In the “Boss model”, together with the relatively high volatility, this is considered as “value destruction” and the “fair value” of Munich Re if they continue like that is something like 15 EUR.
As I have mentioned, this should be not considered to be a hard price target. As Winter pointed out, one could also argue for some general “mean reversion”, however this is not the goal of the exercise.
I am looking for boring companies with a historical good track records of value creation and Munich Re is definitely not such a company.
Greek Stocks: Example Jumbo SA
Some Greek stocks score incredibly well, for instance Jumbo SA
Despite the relative decline in 2011, the company still scores extremely well in the model, as to a certain extent the 10 year average would imply some kind of “mean reversion”. In any other country, Jumbo might be just the boring stock I would be looking for, but being a Greek stock, it is more like the “hot potato stock” (TM rueckzugsgut) I want to avoid.
In the case of Greece, one has to judge potential future developments. So any Greek investment would be definitely a “special situation” and not part of my “Boring sexy stock” universe.
Summary: As expected, the score should not be used for automated investment decision but adds some interesting historical information to the “standard” set of value indicators. The three examples (Coca Cola, Munich Re and Jumbo) show that results seem to be within expectations but should be taken as a starting point only.
At the moment, I am still building up the database, I hope I will be able to show some more comprehensive tables later this week.
thanks for the comment (and checking the model). Tesco indeed looks great in the model. Interestingly, companies with a very strong competitive positon seem to score really well and receive low discount rates. Copmpared to similar stable companies, the P/B multiple for Tesco looks indeed attractive.
Fiskars: It is clear that the model does not capture each and every company. A good exaample is EVN as well. The Verbund stake introduces significant volatality into EVNs equity and the score for EVN is relatively low, despite Verbund itself performing really really well. So yes, the model does not work well for companies holding large stakes in public listed companies which are marked to market against equity.
1. Thanks, was a mistake
2. No the model does not include growth. It assumes a 100% payout of a constant net income based on average ROE (ewige Rente).
Thanks for clarifying the issues.
Rechecking the Coca Cola numbers I got a considerably higher standard deviation using the normal calculation forumla. I assume you used a more sophisticated way to calculate it. One could argue that the positive deviations are no risk at all but positive surprises and eliminate them in the calculation.
Once more a very interesting post.
I’ve just tested your model for two boring looking companies: Fiskars and Tesco.
Fiskars is a Finnish consumer goods manufacturer who concentrates on gardening and household products. Furthermore, it holds a 15% stake in Wärtsilä – a global supplier of marine and energy systems.
This holding is quite interesting for the valuation, because it delivers significant and volatile earnings to Fiskars. Using your model, I calculated a 10-y-BOSS-Score of -0.05, so the Fiskars share looks fair valued, currently. But should one really take the standard deviation of the earnings of an associated company to value the stockholding company? I guess not. One could weigh the earnings contribution of both parts the core business and the holding (here some 50/50). The core business is much less volatile, so I weren’t surprised to find some hidden value.
Tesco, the UK-retailer, shows a fantastic 10-year average TROE of 20.9%. The std. dev. is just 5.5% resulting in a discount rate of 4.85%. This looks suspicious low and the fair value would be more then three times current market cap. (Tesco has just issued bonds with a 5% coupon).
On the other hand, the current market price reflects a future drop of ROE to 10%. This looks a bit too pessimistic, doesn’t it?
PS: Two minor issues (don’t take me oversubtle):
1. In your Coca Cola example, shouldn’t it read ‘Cost of Equity’?
2. You wrote “However by design, any future growth is not included.”
Did you mean any *additional* growth above the calculated average TROE?
Why do you regard every greek stock as hot potato even if exports are 90% of revenue or even if foreign subsidiaries are worth more than actual market capitalization?
Even if the shareholders are expropriated they should still get the assets which are not located in Greece.
I am not an expropriation professional, but as long as I find cheap stocks outside greece I would prefer those stocks.
No one knows what is going to happen.
In my opinion there will be a lot of time to buy cheap Greece stocks in the future. Maybe you miss out the first 20-50%, but if it really turns then they are still cheap.
I would wait until Greece is not headline news anymore. and then 2 years more. then you are suer that the potato is ready to eat without burning your mouth 😉
Der EPS-Sprung 2010 hatte was mit Übernahme von Coca-Cola Enterprises zu tun
klar kann man mit dem Modell nicht jeden Sondereffekt abfangen. Das ist auch gar nicht der Anspruch.
Bei Cocal Cola sehe ich zumindest in den Bloomberg zahlen keinen Sonderfaktor.
Ein einmalig hohes Ergebnis erhöht die Volatilität. Die Volatilität erhöht wiederum den Diskontierungsfaktor signifikant.
Das ist ja das Problem von KGV 10 und Co, der “boss Score” ist hier deutlich besser.
2010 gabs bei CocaCola ein nicht-operativen Sonderfaktor. Deswegen ein deutlich höheres EPS. Das verzerrt deine Durchschnittsberechnung etwas…
Vor allem wenn sowas 2-3 Mal in einem Jahrezehnt vorkommt und man es in die Zukunft hochrechnet.
Ich glaube, dass Siemens um 2000 zweimal mit den Epcos- und Infineon-Börsengängen auch sowas hatte. Da wäre man dann bei einer Hochrechnung wohl voll daneben gelegn
I would get slightly different numbers, 17% ROE, 7.8% std. deviation
Book Multiplier 2.73
2.73/2.2-1 = 0.24
So slightly undervalued according to the model.
Interesting to see that they survived 2001-2003 more or less intact.
Edit: The Citi exapmle clearly shows that this is a backward looking model. No surprise here.
Fast backward some years, I crunched the numbers on Citigroup, looking at it in May 2007:
Average 10y ROE: 18.4%
Std. Dev.: 6%
Fund. Sharpe: 3.1
Value: 845 (per current, heavily diluted share)
Market: ca. 2.2x book = 532/share
There is a problem with the format of the tables, the rightmost columns are not visible.
And: Funtroately, It deosn’t mttaer in waht oredr the ltteers in a wrod are, the olny iprmoetnt tihng is taht frist and lsat ltteer is at the rghit pclae. 😉