Lots of work has been done trying to reverse engineer how Buffett managed to build his $350B behemoth. Tons of books discuss about value investing as practiced by Buffett, while some of the smarter ones delve deeper into the insurance operations that Berkshire operated. When we look closely at the insurance operations, some of the key advantages of this structure are:
- Float generated by insurance offers good amount of leverage for the portfolio that Buffett manages
- Cost of float has been less than zero; Buffett gets paid to manage the float; (see below table) Buffett can borrow cheaper than the U.S. government.
- Float is not directly connected to the markets but to events; the debt will not disappear overnight or cannot be called overnight and hence almost non-recourse
- As long as the insurance companies continues operations, a significant piece of this float will be available for long time (almost permanent capital)
- Insurance operations are exempt from the investment act of 1940 which place restrictions on operations; Berkshire could not have intervened in Geico, Salmon, special situations or own whole public companies; Berkshire would have been regulated as a mutual fund
- Insurance operations can buy whole public companies
.Table from: Paper on Buffett’s Alpha (here)
So, float provided a cheap source of leverage and (almost) permanent capital. What would Berkshire look like if Buffett had invested in S&P 500 instead of handpicking companies? Joseph Taussig of Taussig capital did do the work and reverse engineered the investments.
S&P grew 9.7% CAGR for 40 years between 1969 and 2009. Warren Buffett’s investments grew at 12.4% CAGR. If one had taken the taken all the money (close to $70M when Warren Buffett assumed Berkshire in 1969) and put it in S&P 500, one would have $2.5B and with Warren Buffett’s stock picks, one would have $4.3B. Around 2009, Berkshire’s market cap was about $150B. Why the difference? The study from Taussig capital does a great job of explaining this difference in simple terms, for every $ of equity capital, Buffett used 2$ of reserves, with 12% return on investments, Buffett earned 12% of equity plus 2 times 12% on reserves = 36% net of 4% cost of float (2% for every 1$ of leverage / reserves) and ended with a ROE of 32%. Since Berkshire is domiciled in the US, it has to pay taxes, and the post tax returns are 20% CAGR which takes us close to $150B. The difference between $4.3B and $150B (in 2010) came from the insurance structure.
So, a huge piece of the returns came for Berkshire because of the capital structure than from the investing alpha that Buffett created. Warren Buffett did pick the right insurance companies to buy which mattered a lot. Buffett’s big alpha were from picking the right insurance vehicles that have created the massive conglomerate. As long as the insurance operations do fine and returns on incremental investments is decent, Berkshire will do fine in the future as well.
How does an investor starting today structure an operations with a capital structure better than Berkshire? Remember that Buffett also defers taxes by not selling securities and compounding the float from the taxman. Also note that the ROE’s came down from 32% to 20% because of taxes.
Enter David Einhorn and Greenlight Re. David Einhorn started a re-insurer based in Bermuda (which is tax free) and re-directs all the cash to Greenlight capital to invest which creates better economics than Berkshire does. Although, as Warren often points, insurance as an industry runs at a net loss through the cycle. Greenlight Re has been struggling to keep the combined ratio low and there are additional headwinds for a shareholder. Einhorn charges 2% management fee and 20% of profits of every year (with watermark). It is a very lucrative deal for David Einhorn. With earned premiums close to 50% of capital, the portfolio is levered at 0.5X. If S&P returns 10% and David Einhorn invests to meet the market returns of 10%. With the leverage, it will be 15% minus the 2% management fee and 20% incentive fee, it will be worth 10%. The effects of leverage are being eaten up by Einhorn and not flowing to the shareholders. However, if the unearned premium does go up, it might present interesting possibilities as he has returned 18.9% CAGR on investments in Greenlight Capital since inception in 1996. Einhorn’s ability to create alpha plus the fact that the stock is trading at a discount to book value offers additional possibilities if the unearned premiums go up.
For now, we are content just monitoring Greenlight Re and Third Point Re (Daniel Loeb’s version of the same structure) and watch it for more time.
Beowulf Capital 