Seeking Alpha
Recommended for you:
-
dasdasdasdadas
Special situations, deep value, activist investor, research analyst
Profile| Send Message|
( followers)  

Summary

  • Detailed examination of Berkshire Hathaway using a sum-of-the-parts valuation approach shows it has material upside of up to 36%.
  • We examine a variety of valuation methodologies, and we demonstrate definitively why P/E and P/B are insufficient mechanisms to value Berkshire Hathaway.
  • We engage in a comprehensive deconstruction of Berkshire Hathaway's performance and prospects.
  • We levy various discounts to our model to ensure conservatism in our estimates.
  • We discuss how Berkshire's formidable balance sheet and excess cash allow it to invest in higher returning asset classes than other insurers.

Berkshire Hathaway (BRK.A, BRK.B) has elicited much discussion of late. Recent price appreciation since the announcement of its 2013 results has given some investors confidence that the company will resume its 49-year history of outperformance (for more information regarding historical performance, see page 2 of Berkshire Hathaway's 2013 Annual Report here).

But what is Berkshire Hathaway really worth, and how should it be valued? Why has it been consistently able to outperform other conglomerates, as well as other insurance entities?

We decided to find out.

As a diversified multi-sector holding company, ascribing a meaningful valuation to Berkshire requires more a rigorous methodology than would be appropriate for most "plain vanilla" entities.

To our knowledge, Berkshire Hathaway is not covered by sell-side brokers. We believe this is likely due to the fact that there exists only limited trading volume in the shares, making it challenging for brokers to generate substantial commissions trading the stock (certainly, the Class A shares are too large to trade in significant volume). As well, Berkshire's exceedingly strong balance sheet means it is unlikely to raise capital. Finally, Warren Buffett does not strike us as the type to be particularly receptive to M&A sales pitches from investment bankers. This, to us, appears to be the probable reason that the stock is not covered by the sell side. This provides further reason for us to conduct our own valuation of the company.

Executive Summary

We propose that the most appropriate mechanism by which to value Berkshire Hathaway is by conducting a sum-of-the-parts analysis of its various divisions, which trade on different valuation bases.

This shows that Berkshire has upside of up to 36% (or up to 24%, if we apply a conglomerate discount).

We also note that Berkshire possesses a unique, structural capacity to deliver excess returns with its insurance division on account of:

  1. Non-Insurance Assets: Its diversified portfolio of cash generating, majority-owned non-insurance assets, which exhibit low cyclical exposure and which contribute cash at all times
  2. Balance Sheet: Its robust balance sheet and formidable hoard of excess cash, which provides insulation against claims
  3. Underwriting Profit: The fact that it operates with a significant underwriting profit

Taken together, these items permit Berkshire to weight its investment portfolio asset mix heavily to equities, which represent a higher-average returning asset class.

Unlike Berkshire, most insurers do not possess the excess liquidity required to weather volatile equity market cycles, and as such invest primarily in fixed income instruments, or resort to extensive hedging.

Why Is It Undervalued? We believe that Berkshire is not well-understood by the market. We believe that Berkshire's size, complexity, multifaceted nature, and limited disclosures are the likely reason that it is undervalued. Given the long-term nature of Berkshire's management strategy and the fact that Berkshire tends to acquire with cash, we expect that management is unconcerned by the discount valuation and is unlikely to take any acute action to boost the valuation (aside from continuing to deliver excellent results). If the valuation were to decline to an even greater discount (relative to the company's intrinsic value), the company might buy back stock. Management has indicated that it would buy back stock up to a valuation of 1.2x book value.

Let us first examine our some basic methodologies: We will explain why we reject these in favor of a sum-of-the-parts valuation.

Methodology 1: P/E

We believe that P/E is too basic a metric upon which to value Berkshire Hathaway. Let us first ask the question: how much did Berkshire really earn in 2013?

(click to enlarge)

Looking at the statement of comprehensive income, one may be forgiven for asking what it all means. We examine the major items:

Net Earnings: This is the GAAP definition of after-tax earnings. Once we subtract the minority interest from this, it is what we calculate EPS and P/E on the basis of.

Net change in unrealized appreciation of investments: Berkshire's public market portfolio increased in value by over $25 billion in the last year. However, this is not counted in Berkshire's Net Earnings. In fact, Berkshire's earnings -including its portfolio appreciation- last year would have put it at $45 billion (before taxes on the investment appreciation component).

Applicable Income Taxes: Had Berkshire sold its public market shares and realized the additional $25 billion in earnings, it would have borne an additional $9 billion in incremental taxes. At present, the taxes "theoretically owing" upon sale of the securities at current prices are held on the balance sheet as a deferred tax liability.

Other: We will ignore the other items for the moment, as they net out to almost zero.

This means that Berkshire's low P/E ratio of 15.7 (as of close March 25, 2014) is before the appreciation in its portfolio holdings (although note that Berkshire does employ the equity method on investments it has over 20% voting control of). If we included these comprehensive earnings from its investment portfolio (on an after-tax basis), Berkshire trades at 8.5x the comprehensive earnings.

(click to enlarge)

We acknowledge that it would be a misrepresentation to imply that Berkshire trades at 8.5x earnings. The investment portfolio is volatile: like the stock market. This portion of the business fluctuates, and generates losses in down years for the market. Note 2008, below:

(click to enlarge)

However, by the same token, while we cannot say that gains like 2013 will be the norm, it would also be unfairly penalizing to say that we expect no appreciation in the investment portfolio on average. Although we do not incorporate this into our valuation, we believe it would be fair to ascribe at least an average market return assumption of 6% per annum to the portfolio. This is despite our expectation that Berkshire's portfolio will continue to deliver alpha.

Overall P/E Methodology Results:

(click to enlarge)

The P/E methodology, which grosses up Berkshire's earnings to the average S&P earnings, yields a suggested share price of $243,540 per Class A share and $156.24 per Class B share. This implies a 26% return to a target, based on the market P/E methodology. This uses a definition of earnings which excludes unrealized appreciation in Berkshire's equity portfolio.

Why We Reject P/E: We do, in fact, use P/E to evaluate some components within Berkshire, but we do not use it for the whole business. The numbers yielded by the P/E approach appear to make sense, but we reject this methodology as superficial: Berkshire does not produce widgets - it is a large and complex diversified business with exposure to a variety of sectors, which are each characterized by their own set of business drivers.

Methodology 2: EV/EBITDA (and its derivatives)

Warren Buffett dislikes EBITDA as a gauge of profitability. Presumably this is because sustainable businesses require continuing capital expenditure, and thus, will inevitably have depreciation expenses (as virtually all capital items other than land depreciate). Furthermore, most growing businesses require continuing investments in working capital (some negative working capital businesses do exist, but they are the exceptions).

For the sake of clarity: when we say "EV/EBITDA," we merely mean to imply a capital structure-neutral mechanism, which accounts for leverage and excess cash or other assets. In the case of Berkshire, we refer to the massive cash hoard and investment portfolio, which is far in excess of what is required to manage insurance payouts.

Let us examine Berkshire's cash and debt position, excluding the investment portfolio, which we will address later:

(click to enlarge)

Source: Company Reports

Thus, when we evaluate the use of EV/EBITDA, we include related metrics such as EV/EBIT, and EV/(EBITDA less capex less investment in working capital).

Why We Reject EV/EBITDA: We do, in fact, use EV/EBITDA to evaluate some components within Berkshire, but we do not use it for the whole business. There are 2 main reasons we reject EV/EBITDA (and its derivatives) as a comprehensive valuation metric for Berkshire Hathaway.

Reason 1: Despite the capital-structure neutrality of these mechanisms, they are fundamentally similar to P/E, and fail for the same reason: Berkshire is not a homogeneous entity. The divisions exhibit more differences than similarities. Some would be evaluated on a cash flow or earnings metric, and others would be evaluated on a balance sheet basis.

Reason 2: Berkshire's insurance and the financial divisions do not really possess an "EBITDA," or an "enterprise value," so to speak:

Financial businesses are inherently leveraged: for example, banks possess a leverage ratio. Insurance companies are even more complex, with earnings being based on actuarial assumptions.

Non-financial operating businesses use debt a source of capital, but it is a structural component of financial businesses. By the same token, financial businesses use cash and short-term securities as part of their operations, so these do not necessarily represent excess assets which add to the equity value.

Sum-Of-The-Parts:

We will examine Berkshire Hathaway as a whole, and then look at the divisions individually.

(click to enlarge)

(click to enlarge)

(click to enlarge)

(click to enlarge)

Source: Company Reports

Berkshire's Operating Business Segments

In the filings, Berkshire's operating businesses are broken up into several groups:

Railroads: Burlington Northern Santa Fe is the only component of this.

Energy: This is also referred to as Utilities or MidAmerican, although MidAmerican is, in fact, the name of one of the component assets.

Manufacturing, Services and Retailing: This is the largest division, both by revenue and by EBT, but unfortunately, it is comprised of a relatively broad variety of different business types, many of which are sensitive to a variety of different business drivers.

Finance and Financial Products: This is only a partially accurate description of this division, which is also diversified and which contains some peripheral services, which are not strictly speaking financial in nature.

Together, these comprise what we refer to as the "operating businesses" within Berkshire Hathaway. Here, we show the revenue and EBT by segment.

(click to enlarge)

Source: Company Reports

(click to enlarge)

Source: Company Reports

Railroads: Burlington Northern Santa Fe

Here, we present a comp table with peers of Burlington Northern Santa Fe (BNSF). We compare these to BNSF's operating data, which is how we will derive a valuation.

(click to enlarge)

Comps: UNP, NSC, CSX, KSU, GWR

Note: All prices as of close March 26, 2014

To obtain any insight from this, we "back into" a valuation for this division, based on what we know about BNSF. We have attempted this below.

P/E And Leverage Adjustment: Note that a simple P/E is not appropriate, because it does not take leverage into account: The average P/E of these companies implies a certain market cap for BNSF, assuming a certain leverage ratio.

Thus, we take BNSF's earnings and multiply it by the average P/E ratio of the peer companies. We then make a valuation adjustment to account for the difference between BNSF's leverage and the average leverage within the peer group.

EV/EBITDA: We next take a capital structure-neutral valuation mechanism based on an EV/EBITDA to assess valuation on a different basis.

(click to enlarge)

We then take the average of these two methods to get both an equity value and an enterprise value for BNSF, given the amount of debt and cash attributed to the segment.

P/Book: We ignore P/Book, because railroads are cash flow businesses which do not trade on multiples of book value. Book value is simply not an instructive mechanism by which to assess their value. Moreover, the book value of BNSF will be misrepresentative due to the fact that Berkshire bought the business several years ago, and there will thus be a buyout premium remaining as goodwill on Berkshire's Railroad division balance sheet.

For those interested in the P/Book multiples of the railroad comps, we present them here:

Source: Company Reports, Equity Market Data

Other Valuation Drivers: BNSF's 2014 capex of $3,919 is far in excess of its depreciation of $1,655. This indicates heavy net investment was made in the business recently, and suggests that Berkshire is expecting the division to grow. Large capital investments tend to imply the expectation of an escalation of business levels. Berkshire provides only the vaguest of guidance, so we will not ascribe a valuation premium to BNSF. However, material upside does exist, and we would not be surprised to see BNSF outperform its peers on an operational level in the next several years. We expect that both revenue growth and earnings growth in the division has the capacity to outpace the industry.

We use the blended value of methods 1 and 2 to derive our valuation.

Manufacturing, Service And Retailing

Only Marmon and McLane are named, and have their revenue and EBT provided. We are left to guess at the specifics of the remaining entities, spanning companies (and in some cases brands) such as Dairy Queen, NetJets, Lubrizol, Benjamin Moore, Fruit of the Loom, and numerous others.

(click to enlarge)

Source: Company Reports

Here, we present a comp table with peers of Berkshire's Manufacturing, Service and Retailing divisions.

We acknowledge that this is a somewhat imperfect collection of peers, but the arrangement is not haphazard. It is important to remember that this is a very diverse group. We have sought to provide a collection of category-leading companies which compete with known Berkshire-owned companies.

It is not ideal, but we believe it provides a fairly accurate indication of where this division would trade if it were its own entity.

(click to enlarge)

Comps: SYY, WEN, YUM, MCD, GWW, MSM, WSO, WCC

We have excluded companies such as Fastenal and Chipotle due to their premium valuations; Wendy's trades at a high P/E (although for a different reason than strong companies such as Fastenal and Chipotle), but it does not affect the average valuations materially, as this is a weighted average calculation and not a simple average. Thus, Wendy's is too small to make a difference.

P/E And Leverage Adjustment: Because Berkshire lumps the manufacturing division with its corporate and insurance division ("Insurance and Other"), we have assumed that there is no cash in this division, and ascribed all of the cash to the insurance operation.

As with the prior divisions, we adjust the P/E for the leverage within this division versus the weighted average of the peers.

EV/EBITDA: As with the prior divisions, we next take a capital structure-neutral valuation mechanism based on an EV/EBITDA to assess valuation on a different basis.

(click to enlarge)

P/Book: Here again, we note that P/Book is an unhelpful mechanism with which to evaluate this segment.

(click to enlarge)

Source: Company Reports, Equity Market Data

The profound discrepancies in P/Book multiples within this division indicates that P/Book is not the basis upon which these entities trade.

Other Valuation Drivers: This group is too diversified for us to present a fully unified picture. Although we believe that this segment contains a number of strong, enduring brands which exhibit lower-than-average sensitivity to economic cycles, this is not true of all the constituents. The majority of these businesses are mature, although most exhibit some degree of organic growth. The rapid growth of revenues in this division is also due in significant part to acquisitions.

(click to enlarge)

We use the blended value of methods 1 and 2 to derive our valuation.

MidAmerican: Utilities And Energy

Berkshire's Utilities and Energy division owns a collection of assets, from pipelines, hydroelectric power generating capacity, coal mines and processing facilities, power transmission capacity and even a real estate brokerage.

(click to enlarge)

Source: Company Reports

Here, we present a comp table with peers of the Berkshire's MidAmerican energy and utilities division:

(click to enlarge)

Comps: PCG, EIX, DTE, NEE, ED, D

High Leverage: A large portion of Berkshire's energy holdings comprise stable, heavily regulated entities which lend themselves well to leverage. We have attributed all cash in the combined "Railroads, Utilities, and Energy" section of the balance sheet to BNSF. This has the ostensible effect of raising the effective leverage ratio for the energy division. Even if the cash were, in fact, ascribed to this division, the leverage ratio would still be relatively high.

We are not concerned about the debt load of this division. For one thing, these businesses are meeting their financial obligations. In addition, some of the debt was incurred to facilitate the acquisition, NV Energy. We expect the business to delever over time. Moreover, much of this debt is held by MidAmerican and its subsidiaries; it is secured against all of the assets of MidAmerican, not against Berkshire Hathaway. Note that we do not foresee a situation where this detail would have a practical implication.

P/E And Leverage Adjustment: We make the same adjustments to this division that we make with the others to account for leverage.

EV/EBITDA: As with the prior divisions, we next take a capital structure-neutral valuation mechanism based on an EV/EBITDA to assess valuation on a different basis.

(click to enlarge)

P/Book: Here again, we note that P/Book is an unhelpful mechanism with which to evaluate this segment.

(click to enlarge)

Source: Company Reports, Equity Market Data

Other Valuation Drivers: The highly regulated nature of many constituents within the MidAmerican division affords this division some infrastructure-like properties. The stable cash flows and predictable cash contribution arising from many of the constituents are a boon to Berkshire.

The ostensible high leverage and thin equity capitalization of this division is, to a certain extent, an artifact resulting from Berkshire's disclosures. In any event, it is not concerning to us. We believe this division contributes meaningfully to Berkshire's asset diversification.

We use the blended value of methods 1 and 2 to derive our valuation.

Finance And Financial Products

This is Berkshire's smallest operating division. Here, we present a chart showing the breakdown of the division's EBT. We do this because "revenue" is not as instructive a concept in a financial division, albeit one that we suspect contains both traditional revenue components, as well as net interest income-based contributions.

(click to enlarge)

Source: Company Reports

Below, we present a comp table with peers of Berkshire's Finance and Financial Products division. We acknowledge that these do not represent the best comps, as the comps are large diversified lenders, whereas XTRA and CORT are furniture and small equipment leasing businesses. However, we believe the fundamentals of the businesses are similar, they are merely exposed more heavily to different end-markets (mortgages and houses for most large diversified lenders versus capital goods and consumer durables for XTRA and CORT).

Clayton's business model is somewhat different. In essence, it manufactures and then sells the product to its customers, but it also finances the purchase. It provides the customer with the loan required to purchase the product, operating like a builder and lender in one.

(click to enlarge)

Comps: BAC, JPM, BBT, WFC, USB, PNC

P/E And Leverage Adjustment: We do not make leverage adjustments, as financial companies rely on leverage to conduct business, and Berkshire's division does not appear to be overly leveraged.

EV/EBITDA: Financial companies don't have enterprise values or EBITDA in the conventional sense. Financial companies are inherently leveraged. The best proxy for the enterprise value is the market capitalization. To a financial company, cash is not a superfluous asset; cash is an operating asset for a leasing company; revenue (or in this case, net interest income) cannot be generated without deploying cash.

(click to enlarge)

P/Book: Here, we believe that P/Book and P/Tangible book provides an instructive basis upon which to derive a value for this segment.

Other Valuation Drivers: Given the small contribution to Berkshire's total earnings which the Finance and Financial Products division represents, we believe that the prospect that our comparisons and valuation are imperfect is unlikely to have a material effect on our valuation of Berkshire Hathaway as a whole.

We use the blended value of methods 1, 2 and 3 to derive our valuation.

Alpha Central: Berkshire's Insurance Division

Examining the 10-Ks of most insurance companies, one will notice how much of their investment portfolio is comprised of fixed income securities. This is not the case with Berkshire.

(click to enlarge)

Insurance companies typically possess investment portfolios comprised primarily of fixed income securities, because they rely on the yield to pay out expected insurance claims, which are -actuarially speaking- quite predictable. Note that larger-than-expected claims can be reinsured.

Most insurance companies are relatively light on equities in their investment portfolio. Equities provide a higher average return than fixed income instruments, but they are far more volatile on the whole. As such, they represent a poor match for the steady rate of claims, which requires that cash be made available to pay the claims on a regular basis. When insurers do invest more heavily in equities, they use extensive hedging mechanisms to ensure only modest exposure to equity markets.

Berkshire has overcome this model by holding a massive cash hoard, which allows it to safely invest its float heavily into equities. Berkshire now enjoys the benefit of being able to achieve higher average returns than a normal insurer, but it is also capable of weathering any storm in the equity markets, because it has enough available cash to pay any unforeseen claims, and it does so without extensive hedging (it also reinsures excessive risk losses). Moreover, while most insurers rely on the income generated by their fixed income portfolios, additional cash pours into Berkshire's coffers at all times from its portfolio of stable cash-generating operating assets, such as BNSF, MidAmerican, and the manufacturing and finance groups described above.

Berkshire benefits as well from interest payments on its (relatively speaking) small fixed income portfolio and dividends from its equities. But while it benefits from these, it does not rely on them to manage the cash requirements arising from claims coming from the insurance business.

Finally, Berkshire operates with an underwriting profit, which further increases its resilience and allows the insurance business to be profitable as a standalone entity -even independent of investment performance. Combined with prudent reinsuring of large prospective losses, this means Berkshire would be theoretically capable of funding payouts with premiums.

Summation of Critical Alpha-Generation Capacity: In one respect, Berkshire's portfolio is exceedingly low-leverage; it is very well-capitalized for an insurer. In another respect, it uses this low leverage to take risks that other insurers cannot, such as weighting its portfolio heavily to higher-returning asset classes, such as equities.

(click to enlarge)

Source: Company Reports

(click to enlarge)

Comps: ALL, TRV, MMC, AON, PGR, AFL, XL

P/E: We believe P/E is partially instructive to value Berkshire's insurance division, but it is not sufficient. P/E is helpful to evaluate Berkshire's underwriting profits and the income derived from investments, which it records via the equity method. It typically employs the equity method, where it possesses more than 20% of the voting interest.

However, this does not account for the massive surplus of liquidity (cash and securities) possessed by Berkshire, which we will address below. Simply assigning the division based on P/E undervalues it materially, as it possesses so much excess capital.

EV/EBITDA: As with the financial division, this does not apply.

P/Book: In some ways, we believe that P/Book and P/Tangible book provide an instructive basis upon which to derive a value for this segment. We don't have just the insurance division's standalone book value, as Berkshire consolidates Insurance with "other," in this case, manufacturing. So, we make some adjustments.

(click to enlarge)

Unfortunately, as the calculations shown below (in red) indicate, we do not believe this method makes sense as a mechanism by which to value Berkshire's insurance division.

(click to enlarge)

We see here what the valuation would imply. However, most insurance businesses trade at a premium to book on account of their expected earnings-generation potential. We have already afforded Berkshire P/E-based earnings valuation. Thus, due to its formidable cash and security surplus, the segment has a massive book value. It would thus be unfair to apply the same P/Book or P/Tangible book multiple to Berkshire's insurance division, which has such large excesses. The P/Book, and especially P/Tangible book yield an absurd result.

Simply blending these values represents a flawed method of calculating the value of these businesses.

Berkshire Insurance Valuation Comment: Thus, we believe the best way to value the insurance division is itself a sum-of-the-parts mechanism, whereby we

  1. Assign a multiple based on P/E, and then
  2. Add the surplus asset value from the insurance portfolio.

(click to enlarge)

This yields a valuation of $247 billion, which seems appropriate given the $89 billion of excess investments and $43 billion of excess cash, both of which are valued at only 1.0x book value.

Other Valuation Drivers: We also believe that underwriting profit is a "higher-quality" variety of earnings than normal ordinary insurer profits, and should justify a higher valuation. However, for the sake of conservatism, we do not assume this in our model.

Sum-Of-The-Parts And Conclusion

Based on our sum-of-the-parts analysis, Berkshire's equity value should be $418 billion.

(click to enlarge)

This implies a share price of $254k per Class A share or $170 per Class B share.

Even if we assume a 10% conglomerate discount, which we believe to be unjustified based on Berkshire's track record, we get to a price of $231k per Class A share and $154 per Class B share.

(click to enlarge)

At the risk of sounding too clichéd, we will end with a quote by Warren Buffett:

"Price is what you pay, value is what you get"

At these prices, we are confident that we are getting more than we pay for with Berkshire Hathaway.

Source: What Is Berkshire Hathaway Really Worth? A Comprehensive Look