Summary
- Potential for record operating earnings on Friday.
- Operating businesses currently valued at <10x earnings.="" li="">
- Historically cheap on book value, intrinsic value and look-through earnings measurements.
Berkshire (BRK.B, BRK.A) reports 2Q earnings on August 1st, and while I expect record earnings at Berkshire's non-insurance operating businesses given the combination of a huge volume increase at Burlington Northern, increased earnings at Heinz, and the acquisitions of NV Energy, Phillips Specialty Products, and the remainder of Marmon - my reasons for owning the shares are much longer-term in nature. Even as the market has significantly appreciated to above-trend multiples, Berkshire remains historically cheap on a number of different measurements. This discount to history appears unjustified given its expected financial results. Herein, I will briefly discuss three techniques which can be used to value Berkshire Hathaway and triangulate on my thesis that Berkshire is a bargain.
All forward-looking estimates contained herein are my own are subject to human error which inevitably occurs when humans try to predict the future.
Book Value:
Book value is perhaps the most cited of any Berkshire valuation measurement. Over the last 20 years, Berkshire's average price to book multiple has been 1.58x, with a high of 2.23x in 1995 and a low of 1.17x in 2011. Today, based on my book value estimate of $235 billion ($95.50 per Class B share) for the 2nd quarter, Berkshire trades at 1.33x book value. The standard deviation of Berkshire's price-to-book ratio is .28x, which assuming a normal distribution or "bell curve" would mean that Berkshire trades cheaper than its current level about 19% of the time. Inversely, we would expect Berkshire to trade at a higher multiple 81% of the time.
| Year | P/B | Probability distribution |
| 1985 | 1.50 | 39.33% |
| 1986 | 1.38 | 24.68% |
| 1987 | 1.19 | 8.63% |
| 1988 | 1.58 | 50.04% |
| 1989 | 2.02 | 93.86% |
| 1990 | 1.45 | 32.13% |
| 1991 | 1.41 | 27.12% |
| 1992 | 1.52 | 41.33% |
| 1993 | 1.84 | 82.31% |
| 1994 | 2.02 | 94.02% |
| 1995 | 2.23 | 98.83% |
| 1996 | 1.79 | 77.38% |
| 1997 | 1.80 | 78.53% |
| 1998 | 1.85 | 83.03% |
| 1999 | 1.48 | 35.92% |
| 2000 | 1.76 | 73.16% |
| 2001 | 1.99 | 92.69% |
| 2002 | 1.74 | 71.57% |
| 2003 | 1.67 | 62.24% |
| 2004 | 1.57 | 49.31% |
| 2005 | 1.49 | 38.00% |
| 2006 | 1.57 | 47.96% |
| 2007 | 1.82 | 79.59% |
| 2008 | 1.37 | 23.07% |
| 2009 | 1.17 | 7.74% |
| 2010 | 1.25 | 12.21% |
| 2011 | 1.14 | 6.19% |
| 2012 | 1.16 | 7.16% |
| 2013 | 1.30 | 16.15% |
| Current | 1.33 | 19.06% |
| Mean | 1.58 | |
| St. Deviation | 0.28 |
However, this analysis understates how cheap Berkshire has become. Berkshire has initiated a perpetual and essentially limitless share repurchase authorization at 1.2x book value. Since the repurchase authorization has been enacted, Berkshire has never traded at a price that would allow them to repurchase their own shares. Therefore, it is unlikely going forward that Berkshire would reach the price-to-book ratios that it did during 1987 (1.19x), 2009 (1.17x), 2011 (1.14x), and 2012 (1.16x) because Berkshire would be repurchasing their own shares above this level. If we exclude these observations from our data set, Berkshire's mean price to book ratio moves from 1.58 to 1.64x and the standard deviation moves from .28 to .25x. Using this data, we would expect Berkshire to trade at a premium to its current 1.33x book value multiple 90% of the time.
| Year | P/B | Probability distribution |
| 1985 | 1.50 | 28.51% |
| 1986 | 1.38 | 14.85% |
| 1988 | 1.58 | 39.90% |
| 1989 | 2.02 | 93.50% |
| 1990 | 1.45 | 21.49% |
| 1991 | 1.41 | 16.95% |
| 1992 | 1.52 | 30.55% |
| 1993 | 1.84 | 79.02% |
| 1994 | 2.02 | 93.69% |
| 1995 | 2.23 | 99.04% |
| 1996 | 1.79 | 72.75% |
| 1997 | 1.80 | 74.21% |
| 1998 | 1.85 | 79.94% |
| 1999 | 1.48 | 25.12% |
| 2000 | 1.76 | 67.43% |
| 2001 | 1.99 | 92.08% |
| 2002 | 1.74 | 65.44% |
| 2003 | 1.67 | 54.01% |
| 2004 | 1.57 | 39.09% |
| 2005 | 1.49 | 27.17% |
| 2006 | 1.57 | 37.61% |
| 2007 | 1.82 | 75.55% |
| 2008 | 1.37 | 13.51% |
| 2010 | 1.25 | 5.55% |
| 2013 | 1.30 | 8.20% |
| Current | 1.33 | 10.35% |
| Mean | 1.64 | |
| St. Deviation | 0.25 |
While I find this analysis to be helpful in regard to how Berkshire trades relative to its historical valuation, it is overly mechanical and does not answer the underlying question of whether Berkshire deserves to trade at a discount based on fundamental reasons.
Intrinsic Value:
Buffett has cited two metrics in his annual letter which he describes as being paramount in determining intrinsic value: earnings per share & cash plus investments per share.
By my estimate, Berkshire's operating businesses will earn $13 billion in after-tax earnings in 2014, consisting of $11.2 billion from operating businesses and $1.8 billion in insurance underwriting income. Readers should note, given the large amount of acquisitions it has made, Berkshire amortizes approximately $1.2 billion in intangible assets each year. This charge counts against earnings but does not impact free cash flow nor represent an actual economic phenomenon. For this reason, I tend to back out this expense and its commensurate tax benefit (~$400 million) and estimate Berkshire's actual earnings power at $13.8 billion. Berkshire has 2.46 billion Class B equivalent shares outstanding - so this equates to earning per Class B share of $5.60.
| After-tax non-insurance operating | 2014 Est. |
| BNSF | 3,940 |
| Mid-American | 1,820 |
| Manufacturing, Service & Retail | 4,347 |
| Finance & Financial Products | 1,041 |
| Total operating earnings | 11,148 |
| Insurance underwriting | FY 2014 Est. |
| GEICO | 1,420 |
| GenRE | 536 |
| BHRG | 483 |
| BHPG | 362 |
| Pre-tax underwriting profit | 2,801 |
| Taxes | 984 |
| Net underwriting profit | 1,817 |
| Total earnings | 12,965 |
| Add: After tax amortization of intangibles | 800 |
| Total earnings | 13,765 |
| Class B equivalent outstanding (Billion) | 2.46 |
| EPS | $ 5.60 |
Next we must calculate Berkshire's cash and investments per share. This is fairly straightforward. Based on market appreciation during the 2nd quarter, Berkshire's equity should have a fair market value of ~$123 billion with $30 billion of bonds, $31 billion of hybrid securities and $33 billion of cash ($53 less a stated $20 billion in minimum cash holdings). From this amount I subtract $5 billion of derivative liabilities, $13 billion of holding company debt, and $12 billion as an estimate for the net present value of deferred tax liabilities from investment appreciation. The sum of all of this is $187 billion, or $76 per Class B share.
| Investments ($ billions) | Q2 2014 Est. |
| Equities | 123 |
| Bonds | 30 |
| Hybrids | 31 |
| Distributable Cash | 33 |
| Total cash plus investments | 217 |
| Less: | |
| Derivative liabilities | (5) |
| Holdco. Debt | (13) |
| PV of deferred tax liability | (12) |
| Total liabilities | (30) |
| Aggregate value | 187 |
| Class B equivalent outstanding (Billion) | 2.46 |
| Value per Share | $ 76.02 |
If we deduct the $76 per share in value derived from investments and cash less certain liabilities from Berkshire's current share price $127 dollars, the implication is we are ascribing $51 per share in value to the operating businesses. Considering these businesses should earn $5.60/share, this works out to an earnings multiple of 9.6x earnings. Hopefully, you can appreciate that this appears to be incredibly cheap. It is extremely difficult to find low quality business trading at 10x earnings in today's market, nevertheless high quality businesses with predictable earnings streams like Mid-American and Burlington Northern - which together will generate about half of Berkshire's operating earnings this year.
The S&P is currently trading around 17x earnings (1970 S&P, $115 Earnings), my feeling is that Berkshire's businesses, if they traded as independent companies would likely be valued at a multiple proximate to the S&P 500. However, I tend to be a bit conservative and apply a multiple of 14x, which is a discount to the long-run historical average multiple for the S&P of 15.6x. Using the methodology we derive a value of $78 ($5.60*14) per Class B in regards to the value of Berkshire's operating businesses.
Hence, if we value the operating businesses at $78/share and the cash and assets less associated liabilities at $76/share, this works out to an aggregate value of $154/share. It makes sense to look back and see if this makes sense in the context of Berkshire's historical price-to-book ratio of 1.58x or 1.64x (excluding observations below 1.2x book). Remember that Berkshire's book value should approximate about $95.50 per Class B share in the 2nd quarter, implying that Berkshire would be valued in at 1.61x book value using this methodology. This fits squarely between the two measurements (1.58 and 1.64x) used in the discussion of book value. This is not to say we have found a perfectly precise measurement of Berkshire's inherent value, rather that we can be reasonably comfortable that at 1.33x book value we are getting a pretty good discount to fair value even when using conservative assumptions.
Look-Through Earnings:
Berkshire's stated earnings are punished by GAAP accounting because they are unable to consolidate their proportional share of earnings in equity investments. For instance, Berkshire owns 464 million shares of Wells Fargo. If WFC earned $4.20 this year - that is nearly $2 billion attributable to Berkshire's equity stake of which only $650 million will flow through Berkshire's income statement in the form of dividends. For this reason, I like to add up the earnings power of all of Berkshire's investments and add them to the earnings of the operating business to get a picture of underlying earnings power of the entire concern. Using consensus estimates, I calculate Berkshire's equity portfolio will generate $8.4 in earnings, hybrid securities $1.6 billion, and bonds at $1 billion at a 3% yield.
| Investments earnings | 2014 Est. |
| Equities | 8,413 |
| Hybrid securities | 1,573 |
| Bonds | 1,000 |
| Earnings of investment | 10,985 |
| Operating earnings | 13,765 |
| Total earnings power | 24,751 |
| Class B equivalent outstanding (Billion) | 2.46 |
| EPS | $ 10.06 |
| Multiple | 14 |
| Value | $ 140.86 |
| Add: excess cash per share | $ 13.41 |
| Total value per share | $ 154.27 |
Using my methodology I derive earnings of about $11 billion in Berkshire investments, which added to Berkshire's operating businesses, implies total earnings of ~$25 billion. At a 14x multiple, this implies a total value of $141 per share. However, this analysis excludes any value for Berkshire excess cash of $33 billion or $13/share. Obviously the cash does not generate earnings but it is clearly of value given it could be distributed to shareholders or used to purchase additional businesses. Adding the two numbers gets to $154/share, the same valuation we arrived at in the previous section and a price to book ratio of 1.61 which is entirely in-line with historical averages.
Conclusion:
Berkshire's historic price-to-book ratio, when viewed over a long period, probably provides a reasonable basis off of which to calculate the inherent business value of the Company. Currently, Berkshire's historically low price-to-book multiple along with conservative alternative valuation techniques would imply the shares are undervalued by 20 to 25%. It goes without saying that investors should do their own work and analysis before making any investment conclusion.
