Waste management was a 'disaster' apparently - It dropped 70% after Ben Miller purchased the stake, He said "I wish I had put greater weight on cash flow that appeared relative to underlying earnings"
Thursday, June 26, 2014
Friday, June 20, 2014
Kelly's Criterion
fraction of capital to bet in a single bet is p - q/W
where p stands for probability of success, q for failure (1-q) and W is the ratio of amount you win when you win to amount you lose when you lose.
i.e. W = w/l ; where w stands for win amount when you win and l stands for loss amount when you lose - simply put it is the odds
or edge/Odds i.e. (pW-q)/W, which can be rewritten as (pw-ql)/w
applications to stock market:
Yes bank - Aug 2013
CMP - 240
(all this after doing fundamental analysis on the bank, this is the most important edge that sits atop Kelly's edge:))
Assume peak P/ABV multiple is 2.5 and prevailing multiple is 1.2 x; bottom most multiple is 1.0 x
then W can be computed as 2.5 x - 1.2 x divided by 1.2 x - 1.0 x i.e. 1.3:0.2 or 13:2
One way to establish ps and qs is to look at the frequency table of historical pricing multiples
Assume that only 15% of the time, the stock has traded at a P/B less than 1.2 x => p = 1 - 0.15% = 0.85%
Now applying Kelly's f = 0.85% - 0.15%/6.5 = 0.82 % i.e. odds are favorable bet heavily (Charlie Munger)
However, not every man has steel balls and hence it could be morphed to 0.82/2 i.e. 40% of one's investable wealth should have been invested in a single stock.
where p stands for probability of success, q for failure (1-q) and W is the ratio of amount you win when you win to amount you lose when you lose.
i.e. W = w/l ; where w stands for win amount when you win and l stands for loss amount when you lose - simply put it is the odds
or edge/Odds i.e. (pW-q)/W, which can be rewritten as (pw-ql)/w
applications to stock market:
Yes bank - Aug 2013
CMP - 240
(all this after doing fundamental analysis on the bank, this is the most important edge that sits atop Kelly's edge:))
Assume peak P/ABV multiple is 2.5 and prevailing multiple is 1.2 x; bottom most multiple is 1.0 x
then W can be computed as 2.5 x - 1.2 x divided by 1.2 x - 1.0 x i.e. 1.3:0.2 or 13:2
One way to establish ps and qs is to look at the frequency table of historical pricing multiples
Assume that only 15% of the time, the stock has traded at a P/B less than 1.2 x => p = 1 - 0.15% = 0.85%
Now applying Kelly's f = 0.85% - 0.15%/6.5 = 0.82 % i.e. odds are favorable bet heavily (Charlie Munger)
However, not every man has steel balls and hence it could be morphed to 0.82/2 i.e. 40% of one's investable wealth should have been invested in a single stock.
Tuesday, June 17, 2014
Super Read - Source: Valueandopportunity.com
As I did not find historical P/Es for the Dax in 1987/1988, let’s look at this table of historic P/Es for the S&P 500:
| P/E | |
|---|---|
| 31.12.1973 | 12,3 |
| 31.12.1974 | 7,3 |
| 31.12.1975 | 11,7 |
| 31.12.1976 | 11,0 |
| 30.12.1977 | 8,8 |
| 29.12.1978 | 8,3 |
| 31.12.1979 | 7,4 |
| 31.12.1980 | 9,1 |
| 31.12.1981 | 8,1 |
| 31.12.1982 | 10,2 |
| 30.12.1983 | 12,4 |
| 31.12.1984 | 9,9 |
| 31.12.1985 | 13,5 |
| 31.12.1986 | 16,3 |
| 31.12.1987 | 15,6 |
| avg | 10,8 |
Someone like John Hussmann might have said that stocks have nowhere to go as the P/E even after the 1987 crash was ~50% higher than the preceeding 15 year average. At the and of 1987, 10 30 year US Treasuries were yielding around 9%, another argument why stocks didn’t look that “apetizing” at that point in time. Why bother with stocks if you can earn double digits with corporate bonds any time ?
What followed
Looking back, it is easy to point out some of the events which led to this remarkable run especially for the DAX over the last 26 years:
- Communism broke down (“Peace dividend”)
- the Eurozone was created, stimulating cross border trading, increasing competition
- technology change (PC, Internet, Mobile)
- Corporate taxes in Germany went down form >50% to ~30%
- interest rates declined for now 25 years in a row
- old crossholding structure (“Deutschland AG”) dissolved, more professional management, foreign investors
- the BRIC story unfolded, further possibilities to export “core competency” goods like machinery and cars
- the Eurozone was created, stimulating cross border trading, increasing competition
- technology change (PC, Internet, Mobile)
- Corporate taxes in Germany went down form >50% to ~30%
- interest rates declined for now 25 years in a row
- old crossholding structure (“Deutschland AG”) dissolved, more professional management, foreign investors
- the BRIC story unfolded, further possibilities to export “core competency” goods like machinery and cars
In 1987/1988, few market pundits did even predict a single one of those factors. That’s why I think that just looking into the rearview (valuation) mirror should not be the only tool in the investing toolbox. Past P/Es will not predict future seismic shifts. On the other hand, one should not rely on such evcents happening over and over again and boosting share prices further. Clearly, interest rates and taxes will not fall that much lower and the effect of the end of Communism will not repeat itself.
For me the major conclusion is the following: Do not rely on any one system which tries to predict the future and/or future returns. Keep an open eye on everything, from valuations to macro economic factors and political shifts. Be prepared for surprises. Inthe long term, many surprises turned out to be positive for the economy and stock return.
Some musings on the Dax constituents
Just for fun, I created a table with the long term performance of the 15 “surviving” Dax constituents. Unfortunately I only got performance numbers back to 1992, but the p.a. Performance of the DAX was quite similar. lets look at those 15:
| 1987 | Still in DAX | Comment | LT Perf (08/1992) | p.a. |
|---|---|---|---|---|
| DAX | 545,14% | 8,95% | ||
| Allianz * | 1 | 177,55% | 4,80% | |
| BASF * | 1 | 3650,23% | 18,12% | |
| Bayer * | 1 | 1598,15% | 13,90% | |
| BMW * | 1 | 1723,82% | 14,28% | |
| Commerzbank * | 1 | -70,14% | -5,40% | |
| Continental | 1 | 1962,28% | 14,92% | |
| Daimler-Benz (*) | 1 | 90,50% | 4,22% | |
| Deutsche Bank * | 1 | 89,57% | 2,98% | |
| Deutsche Lufthansa * | 1 | 615,84% | 9,47% | |
| Henkel * | 1 | 1200,08% | 12,51% | |
| Linde * | 1 | 699,66% | 10,03% | |
| RWE * | 1 | 308,71% | 6,68% | |
| Siemens * | 1 | 742,92% | 10,29% | |
| Thyssen (*) | 1 | 89,98% | 4,32% | |
| Volkswagen * | 1 | 1690,10% | 14,18% |
Not surprisingly, financial stocks do not look good here. Overall, companies which are considered “well managed” did quite well such as Henkel, Bayer, BMW, Linde. Surprising for me is the fact that Lufthansa actually outperformed the DAX as well as Siemens.
Now let’s take a quick look at the new stocks. If I didn’t have returns from 1992, I made a comment:
| Total | p.a. | Perf. Since | |||
|---|---|---|---|---|---|
| Adidas | 1 | 896,84% | 13,23% | 1995 | |
| Beiersdorf | 1 | 1658,99% | 14,09% | ||
| Deutsche Börse | 1 | 335,79% | 11,74% | 2001 | |
| Deutsche Post | 1 | 103,80% | 5,41% | 2000 | |
| Deutsche Telekom | 1 | 62,22% | 2,80% | 1996 | |
| EON | 1 | 485,63% | 8,46% | ||
| Fresenius | 1 | 4651,42% | 19,42% | ||
| Fresenius Medical Care | 1 | 174,05% | 5,90% | 1996 | |
| HeidelCement | 1 | 242,80% | 5,83% | ||
| Infineon | 1 | -80,66% | -10,95% | 2000 | |
| K&S | 1 | 3084,30% | 17,24% | ||
| Lanxess | 1 | 302,98% | 16,11% | ||
| Merck | 1 | 555,53% | 10,64% | 1995 | |
| Munich Re | 1 | 300,42% | 7,24% | 1994 | |
| SAP | 1 | 3502,32% | 19,98% |
Not surprisingly, the best “newcomers” also lead the total Dax performance. Smaller companies which grow big are always the best investments, although it is often hard to identify them before.
Finally one other table. Let’s look at some of the best performers and their historical P/Es:
| FRE | SAP | HEN3 | BEI | BAS | |
|---|---|---|---|---|---|
| 31.12.1992 | 28,6 | 24,4 | 19,6 | 18,9 | 11,4 |
| 31.12.1993 | 35,2 | 25,8 | 25,7 | 22,8 | 28,0 |
| 30.12.1994 | 19,4 | 36,7 | 15,0 | 20,6 | 14,6 |
| 29.12.1995 | 33,0 | 55,2 | 18,4 | 18,9 | 7,8 |
| 31.12.1996 | 64,4 | 52,1 | 25,9 | 28,7 | 14,4 |
| 30.12.1997 | 49,2 | 61,1 | 29,5 | 46,8 | 12,0 |
| 30.12.1998 | 30,8 | 71,5 | 32,8 | 30,4 | 11,8 |
| 30.12.1999 | 27,1 | 83,7 | 26,2 | 32,5 | 25,3 |
| 29.12.2000 | 37,7 | 60,0 | 21,7 | 41,9 | 23,6 |
| 28.12.2001 | 183,3 | 78,5 | 18,3 | 38,1 | 20,7 |
| 30.12.2002 | 10,8 | 46,3 | 20,0 | 31,3 | 13,9 |
| 30.12.2003 | 23,0 | 38,1 | 17,1 | 27,3 | 27,5 |
| 30.12.2004 | 18,2 | 30,9 | 5,3 | 21,9 | 14,5 |
| 30.12.2005 | 20,1 | 31,5 | 16,2 | 23,7 | 11,3 |
| 29.12.2006 | 23,5 | 26,2 | 18,9 | 16,7 | 11,6 |
| 28.12.2007 | 21,5 | 22,3 | 18,1 | 27,2 | 12,1 |
| 30.12.2008 | 21,2 | 15,5 | 54,7 | 16,8 | 8,9 |
| 30.12.2009 | 14,2 | 22,3 | 26,4 | 27,8 | 28,2 |
| 30.12.2010 | 16,3 | 24,9 | 18,1 | 29,7 | 12,0 |
| 30.12.2011 | 16,9 | 14,0 | 16,7 | 39,8 | 8,0 |
| 28.12.2012 | 16,3 | 25,8 | 18,3 | 31,6 | 13,6 |
| 30.12.2013 | 19,7 | 22,3 | 23,1 | 31,3 | 14,7 |
We can easily see that quality and growth NEVER is cheap. I am not sure if that Henkel 2004 P/E of 5 is incorrect data, but the solid “quality stocks” always traded “richly” and nevertheless delivered outstanding long term performance. Only BASF, as a “quality cyclical” company has been available at single digit P/Es at some years.
So after all, this is wat Warren B. likes to tell us: In the long term, quality does seem to beat anything else, especially if you factor in taxes, trading costs etc.
Summary:
So what does this all tell us ? I am afraid that I cannot come up with some “Magic Formula” to identify future winners. Nevertheless, I think the look back emphasizes three of Warren Buffet’s main points:
1) over the long term, stocks have been a unbeatable compounding machine. A return 10 times the original inevstment in 26 years despite several devasting crashes speaks for itself
2) over such a long time horizon, it seems that “quality buy and hold” seems to be at an advantage at least for large caps. Yes, introducing a backtested system (market timing, EV/anything) could generate fantastic returns as well, but just buying and holding well managed companies did produce spectacular returns
3) Just buying the index and sitting on one’s ass would have beaten almost all active strategies. To be fair although, the first DAX index funds were available mid/end 90ties…..
Subscribe to:
Posts (Atom)