The Big Idea
~
We are big fans of Warren Buffett and all that he has achieved. This point should be kept in mind at all times as you read the content on this website. However, we feel that far too much of his success is attributed to his stock picking, rather than his use of the insurance and reinsurance businesses for his outsized success.
Buffett was a hedge fund manager for 13 years. He quit cold turkey to go into insurance, reinsurance, and banking, because he deeply believed he could deliver better returns if he managed the assets of an insurer, reinsurer, or bank than he could if he only managed assets for a fund. In 1979, the U.S. government gave him the choice of exiting the insurance and reinsurance businesses or the bank and he sold Illinois National Bank, leaving insurance and reinsurance as his vehicles for outperformance.
In comparison to returns from a fund or managed account, an insurer, reinsurer, or bank enjoys two advantages. The first advantage is a form of leverage. In the case of insurers and reinsurers, Buffett calls this “float”. In the case of banks, we are referring to government guaranteed term deposits of greater than one year which can be readily rolled over as opposed to demand deposits that are used for paying bills and accessing ATM machines (which can be withdrawn on a moment’s notice) or debt that is often difficult to roll over.
The differences between float or term deposits and demand deposits or debt leverage are significant. In an operating company, leverage is converted to assets such as plant and equipment, inventories, and receivables that may not be liquid enough to repay the leverage when it is due and may not be rolled over. In the case of financial assets, leverage usually means a demand loan or deposit, the terms and availability of which can change on a moment’s notice (i.e. Bear Stearns, Lehman Brothers, and Northern Rock in the crisis).
Buffett goes to great lengths to disparage leverage and distinguish it from float. Float and term deposits are stable in cost and duration, relatively predictable, and do not require disorderly liquidations of assets to meet an obligation.
In the case of insurers and reinsurers, the industry-wide cost of float is roughly 3% per year. To the extent that the insurer or reinsurer invests the float at a higher rate of return than its cost, it earns a spread (similar to the spread that a bank earns if its interest income, net of loan losses, exceeds its deposit costs). Thus, pre-tax ROEs for insurers and reinsurers are a function of the investment returns on assets plus the number of turns of leverage times the spread per turn.
Buffett has always used a moderate level of float. In the early days, Berkshire used roughly $2 of float for $1 of equity capital. Today, Berkshire uses about 60 cents of float for each $1 of equity capital. By comparison, Swiss Re is levered 6 times and Generali’s leverage to tangible book value has been as high as 40 times.
One of the great myths about Buffett is that he has always gotten float for less than nothing (in other words he is paid for holding the float). Until he bought Geico, Buffett never consistently achieved a negative cost of float and until 2006 (the year after Katrina), he never cumulatively had a negative cost of float. According to a paper by AQR, Berkshire’s cumulative cost of float was roughly 2.2% per year for the first 40 years (still better than the average), but he invested the proceeds at rates of return much higher than 2.2% per year.
If an asset manager invests the equity capital of an insurer, reinsurer, or bank as he or she would a fund or managed account, and the financial institution breaks even operationally, the financial institution and the fund would have identical pre-tax returns. However, non-U.S. insurers, reinsurers, and banks pay little or no income taxes on annual earnings. Thus, after-tax ROEs of a non-U.S. financial institution should be greater than a fund with an identical investment strategy, when the financial institution breaks even operationally and the earnings are 100% derived from returns on an identical amount of equity capital.
To the extent that assets earn more than the cost of float or deposits less operating expenses, the pre-tax ROEs for the financial institution will be higher than those of the fund. Since neither the financial institution nor its shareholders are taxed annually on its earnings, this difference is even greater for investors who are annually taxable on investment income and realized gains.
We call this Big Idea of using an insurer, reinsurer, or bank rather than a fund or a managed account to deliver asset management skills “Structural Alpha” (the reinsurance versions of this are often referred to as "total return reinsurers").
Aside from Buffett, more than 60 asset managers and direct lending funds (or variations) have either acquired or started insurers, reinsurers, or banks, where they manage all of the assets of the institution and usually do so for full fees. Some of the better known asset managers who have done this include:
Apollo
AQR
Blackrock
Blackstone
Cerberus
Citadel
D.E. Shaw
Goldman Sachs
Greenlight
Harbinger
HBK
Highbridge (twice)
Knighthead
Magnetar
Mariner
Maverick
Moore
Paulson
SAC
Soros (4 times)
Third Point
Two Sigma
We believe that we have advised more asset managers who have successfully launched or acquired an insurer, reinsurer, or bank in order to manage its assets than all the other asset managers combined who have successfully launched or acquired an insurer, reinsurer, or bank without our advice. Only three of our insurers, reinsurers, or banks has ever underperformed the manager’s funds for their lifetimes or their first 10 years (whichever was shorter), and those managers had lost money as fund managers (leverage works both ways).
As such, we also believe that any investment strategy that can deliver pre-tax returns in excess of 5% per year over any 5 year rolling period of time can replicate the Buffett model and outperform his or her funds or managed accounts with a high degree of certainty.
The second advantage of Structural Alpha is a premium to book value if the insurer, reinsurer, or bank is publicly traded. Logically, if a tax efficient insurer, reinsurer or bank that breaks even operationally delivers identical pre-tax returns as an open ended fund, it should be valued higher than the open ended fund because of its tax efficiency.
Since open ended funds transact at NAV (book value), it stands to reason that a publicly traded, tax efficient insurer, reinsurer, or bank should trade at a premium to book value. At the end of 2013, Berkshire and Third Point Re traded at 1.4 times book value on the New York Stock Exchange and Greenlight Capital Re traded at 1.3 times book value on NASDAQ.
By comparison, a closed ended fund or a Business Development Corporation ("BDC") (other forms of permanent capital) are virtually certain to trade at a discount to NAV. If an investor can readily enter or exit an open ended fund at NAV, no one would pay more than NAV to enter a closed ended fund or BDC and would only buy into a closed ended fund or BDC at a discount to NAV, because of the uncertainty of exiting at NAV as one can in an open ended fund.
Only five asset manager backed reinsurers have gone public, and we advised four of the five. Three of them (Scottish Re, sponsored by Maverick, Max Re, sponsored by Moore, and Resource Re, sponsored by Kudu) have been acquired or merged and are no longer independent. However, the two remaining publicly traded asset manager backed reinsurers, Greenlight Capital Re (NASDAQ: GLRE) and Third Point Re (NYSE: TPRE) provide a treasure trove worth of publicly available information.
In 2004, David Einhorn seeded GLRE with $50 million. Had he left the $50 million in his fund, his wealth would have increased by $47 million over the next 10 years. Had his non-U.S. investors left $50 million in his offshore funds instead of investing in GLRE, their wealth would have increased by $60 million over the same 10 years. Instead, in GLRE, wealth had increased by $115 million for the offshore investors and Einhorn alike. During that same 10 year period of time, GLRE outperformed Berkshire Hathaway by 6.5% per year.
Since insurers, reinsurers, and banks are exempt from Passive Foreign Investment Company ("PFIC") taxation, there were no K-1s nor annual taxes on GLRE’s earnings for U.S. investors and taxable investors in jurisdictions such as Australia, Canada, the UK and the U.S. were only taxed at capital gains rates if and when they sold their shares instead of ordinary income rates when they redeemed shares in an open ended fund. In addition, GLRE provided roughly $3 million of daily liquidity on NASDAQ, whereas the Greenlight fund investors were subject to lockups, periodic redemption and notice periods, and could be gated.
To neutralize the insinuation that hedge fund reinsurers were tax dodges, all of the initial capital in Greenlight Re, except David Einhorn’s, was raised from non-U.S. investors, and Einhorn never got a tax benefit until Greenlight Re was no longer a PFIC.
In this case, Structural Alpha had been worth 4.7% per year for non-U.S. investors for nearly 10 years. If you look at the difference between Greenlight Capital’s offshore funds and its U.S. funds, the difference is 1.8%, which we call Tax Alpha. For a U.S. investor, GLRE’s Structural Alpha has been nearly 3 times its Tax Alpha and tax is a tertiary (maybe even a fourth level) benefit relative to higher ROEs and a premium to book value (for some investors, daily liquidity would be the third level benefit).
On this last point, tax authorities take the view that activities undertaken solely to reduce taxes are abusive and that the tax position should be challenged, whereas additional tax benefits for investments that are made for economic reasons irrespective of taxation are generally accepted by tax authorities.
This significant outperformance does not take into account the considerable fees that GLRE generated for Einhorn, which could be tax deferred through the end of 2008, where they could then compound tax deferred through the end of 2017. As of the end of 2018, GLRE had generated $368 million in asset management and performance fees. By comparison, Dan Loeb seeded Third Point Re with $75 million and has earned $465 million in fees.
The following pieces track the transition of Buffett from hedge fund manager to the master of Structural Alpha:
Carol Loomis on Hedge Funds (1966) – The editor of Buffett’s annual letter calls him a hedge fund manager
Carol Loomis on Hedge Funds (1970) – She again refers to Buffett as a hedge fund manager
Buffett’s Alpha – AQR’s 2012 paper on Buffett
The Economist’s Review of AQR’s paper on Buffett’s Alpha
Professor Sanjay Bakshi’s presentation on Buffett’s success - https://www.slideshare.net/bakshi1/floats-and-moats
Buffett’s Secret Sauce – Forbes.com weighs in
Buffett on the Advantages of Float – In his own words
Those who are interested learning more about this Big Idea may also download the following:
A Tale of Two Capital Structures – Our 2010 white paper on Structural Alpha
A December 2013 piece by E&Y on the subject of hedge fund backed reinsurers
IRS Notice 2003-34 – The IRS questions the PFIC exemption for hedge fund reinsurers in 2003.