How We Value Berkshire Hathaway
Alternative methods have some use, but we think discounted cash flow is the most fundamentally sound way to value the conglomerate, as we discuss in the final installment of our 5-part series.
Ahead of Saturday's Berkshire Hathaway (BRK.A) (BRK.B) Annual Meeting we're taking a closer look at the best way to value the complex company. We believe that understanding the benefits and shortfalls of different methodologies can provide valuable insight into the ways which different investors are approaching the firm's overall valuation.
Part 1 of the series, on an earnings-based multiple approach, is available here. Part 2 on book value can be found here. Part 3 on the two-column approach is here. Part 4 on using insurance float to value the firm is here. These approaches are useful for triangulating estimates, but are not robust enough by themselves.
The methods we have examined thus far, which are summarized in the table below, provide a wide range of potential intrinsic values for Berkshire.
While some approaches seem to provide values that are fairly close to the market price or our fair value estimate, we believe that all the approaches are flawed or lacking. In our opinion, each has its failings due to either oversimplification, a lack of intuitive justification or a lack of applicability to peer companies. Notably, many of the methods for valuing the insurance business do not seem to work consistently when applied to other insurers, which throws their validity for Berkshire into question. These approaches may simply appear to work for Berkshire by coincidence or jerry-rigged assumptions rather than being a fundamentally sound way to value the company.
A DCF Model that Explicitly Forecasts Sources of Berkshire's Profitability Provides the Best Results
When calculating our fair value estimate for Berkshire Hathaway's shares, we use a sum-of-the-parts methodology that values the different businesses separately and then combines these values to arrive at a total value for the firm. As part of this process, we use discounted cash flow methodologies to value each of the company's major segments: Insurance; Railroad, Utilities & Energy; Manufacturing, Service & Retailing; and, Finance & Financial Products. We believe that our discounted cash flow approach allows us to explicitly model all of the cash flows, along with necessary reinvestments and excess capital, associated with Berkshire's total enterprise, and therefore provides investors with a more robust and reliable valuation of the company's shares than any of the shortcut or alternate methods we've examined.
Our valuation model for Berkshire is built on our insurance DCF template with supplemental models created for the non-insurance businesses in the company's portfolio. Given the segmentation that exists within Berkshire's reported financial statements, as well as the fact that both Burlington Northern Santa Fe and MidAmerican Energy Holdings file quarterly and annual reports with the SEC, we are able to strip out and model the results for these businesses. With our discounted cash flow model for insurance companies, we are able to explicitly forecast the income statement, balance sheet, and resulting cash flows for Berkshire's insurance operations, accounting for earned premium growth, loss and expenses of the insurance operations (which leads to the booking and increases/decreases of reserves), investment cash flows, and the required capital levels of the business, among numerous other factors, during the initial projection period. In contrast to many of the other approaches we've mentioned, we believe this method captures all the different dynamics and moving parts associated with Berkshire's insurance operations.
With regard to the company's noninsurance operations, which encompass a wide array of businesses, including Burlington Northern Santa Fe (railroad), MidAmerican Energy (energy generation and distribution), McLane (food distribution), Marmon (manufacturing), Shaw Industries (carpeting), Benjamin Moore (paint), Fruit of the Loom (apparel), Dairy Queen (restaurant), and See's Candies (food retail), we model four distinct segments--Burlington Northern, MidAmerican, the firm's Manufacturing, Service & Retailing segment, and its Finance & Financial Products division--based on the level of information we have to work with for each of these areas of operation. Much like we do with Berkshire's insurance operations, we use a discounted cash flow approach to model the growth and profitability of each of these different segments, along with the cash flows and investments necessary to support this growth, to arrive at our fair value estimates.
In a few cases, we will triangulate the DCF-derived estimates with multiple-based approaches--including EV/EBITDA and price/book--to arrive at our final per share value estimate. As part of this process, we also run Burlington Northern and MidAmerican, both of which file quarterly and annual reports, through our general DCF model, further ensuring that our fair value estimate for each of these subsidiaries is within range of the stand-alone valuation that can be produced for each firm. Once we've calculated our per share fair value estimates for each of the segments, we roll them up to arrive at our total value for the firm. While we believe that our approach to valuing Berkshire allows us to capture more than a few important valuation factors that a quick back-of-the-envelope approach is likely to miss, we also recognize the fact that, as with most models, a DCF model is only as good as its inputs, which require a broader level of knowledge about each of the operations being valued.
Berkshire Hathaway's Insurance Unit Valuation
We estimate that Berkshire's insurance operations are worth $93,900 per Class A share (or $63 per Class B share). Our forecast for premium growth, while averaging 6%, is driven by an assumption of an eventual hardening of the insurance pricing market. There has been increasingly positive news on this front more recently, and we believe that Berkshire's insurance operations, in particular its reinsurance business, stand to benefit from significant price increases when the pricing gains accelerate. We believe that GEICO will continue to grow premiums at a pace that is slightly above the industry's average rate of growth, as direct-selling continues to take incremental share from the agency channel overall. In the long run, though, we expect these growth rates to slow, as insurance companies tend to grow roughly in proportion with the rate of GDP and inflation.
With regard to total investment return, we believe that Berkshire will continue to have success in its investments, as evidenced by the above average rate of return assumed on the portfolio, which reflects a combination of both investment yield and market appreciation. The higher average rate of return is largely driven by Berkshire's higher allocation to equities, as well as its ability to use the strength of the firm's balance sheet to procure higher yielding investments, all of which have contributed to the higher long-term returns that Berkshire has historically been able to generate through this strategy. And, finally, we assume a 6% average underwriting margin for the firm. While Berkshire's insurance business is subject to volatility through its catastrophe underwriting, we expect this business (on average) to be solidly profitable. It should also be noted that GEICO provides a consistently high level of underwriting profitability, which we believe will continue because of its cost and scale advantages.
As mentioned above, we believe that book value is an appropriate multiple to use for triangulating and rationalizing our fair value estimates for insurance companies. On that basis, the value for Berkshire's insurance operations seems about right at 1.4 times book value, given that the business is about equally split among GEICO and Berkshire's other property-casualty operations and its reinsurance operations, comprised of General Re and BHRG.
Railroad, Utilities & Energy Valuation
Of the more than 70 noninsurance businesses in its portfolio, the two largest contributors to Berkshire's pretax earnings are Burlington Northern and MidAmerican. As both subsidiaries file quarterly and annual reports with the regulators, we are able to create more detailed valuation models for their operations. In aggregate, we estimate that Berkshire's Railroad, Utilities & Energy operations are worth $43,200 per Class A share (or $29 per Class B Share), with more than two thirds of that value coming from Burlington Northern, which continues to generate strong results for Berkshire. BNSF operates one of the largest railroad systems in North America, with around 32,000 route miles of track in 28 states (primarily west of the Mississippi River) and two Canadian provinces. During the last decade, the railroad has generated an 8% revenue CAGR and average operating margins of around 25%. The business has traditionally required at least $2 billion in capital expenditures annually, and carried a fair amount of debt on its books (leaving it with a WACC of around 8%).
In our base case scenario for the railroad operations, we assume that BNSF generates average annual revenue growth of around 5% longer term. While this may seem conservative given the firm's history, as well as the fact that most railroads have seen a significant improvement in pricing power over the past decade (based in part on their ability to pass through fuel price increases), we feel that a 5% CAGR for revenue represents a fairly stable blend of volume growth and pricing over the long run. With regards to profitability, BNSF is likely to close out 2012 with pretax margins of around 28% of revenue. We see future results trending down to around 27%, due to higher operating costs longer term. This leads to a $31,500 per Class A share (or $21 per Class B Share) value for BNSF.
The fair value estimate for BNSF relative to earnings appears to be about right at 15.0 times, relative to 14.5 times for Union Pacific and 14.7 times for Norfolk Southern, and a group average of 15.3 times.
Looking more closely at the utilities and energy assets, Berkshire holds both energy generation (PacifiCorp, MidAmerican Energy Company, and Northern Powergrid) and energy distribution (Northern Natural Gas and Kern River) assets, which are consolidated under MidAmerican Energy Holdings. The majority of the holding company's revenue and profitability (as well as its ongoing capital investments) is driven by its two main regulated utilities--MidAmerican Energy and PacifiCorp. With regulators typically setting customer rates that allow utilities to earn 10%-12% returns on equity, and MEHC being a fairly active acquirer of assets over the years, the holding company generated a 9% revenue CAGR and operating margins of around 23% during the last decade. The business has traditionally required at least $2 billion in capital expenditures annually, and has also carried a fair amount of debt on its books, leaving it with a WACC between 7% and 8%.
In our base-case scenario for MEHC, we assume that the company generates average annual revenue growth of around 4% longer term. While this may seem conservative given the firm's history, results have been much more tepid over the last five years. As such, we feel that 4% average annual revenue growth is a fairly good estimate in an environment where the economy is slowly working its way back to full potential, and MEHC continues to invest in additional capacity. With regards to profitability, the holding company is likely to close out 2012 with operating margins of around 24%. We see future results trending back down to 23% of annual revenue, which would put results in line with MEHC's average profit levels during both the last five- and 10-year periods. We also assume that capital expenditures are somewhat higher than they have been in the past, as MEHC spends more to upgrade its energy generation and distribution networks. This leads to an $11,900 per Class A share (or $8 per Class B Share) value for MEHC, adjusted for Berkshire's 89.8% ownership interest in the firm.
Not unlike what we saw with the insurance segment, book value per share can be used to triangulate and rationalize our fair value estimates for Berkshire's utilities and energy segment. Despite its blend of energy generation and distribution assets, the segment's implied multiple of 1.2 times sits much closer to the multiples for the regulated utilities-- Duke Energy (DUK) and Xcel Energy (XEL)--we've highlighted in our comparison table, which makes sense, given that the majority of the holding company's revenue and profitability is driven by its two main regulated utilities--MidAmerican Energy and PacifiCorp. The implied multiple also sits closer to the regulated utilities, as opposed to the pipeline operators, in the collection of peers--Duke Energy, Xcel Energy, American Electric Power (AEP), Wisconsin Energy (WEC), TransCanada (TRP), and Spectra Energy (SE)--that we defer to when triangulating our own valuation estimate for MEHC. This is also the case when looking at MEHC on a fair value/earnings basis, with the utilities and energy segment's implied multiple of 13.3 times, more in line with Duke Energy and Xcel Energy, the two regulated utilities in our comparison table, and being well below the group average of 15.3 times (but just slightly below the average of the regulated utilities of 13.9 times).
Manufacturing, Service & Retailing Valuation
Berkshire's manufacturing, service and retailing operations are the next largest contributor to the firm's overall value and include a wide array of businesses operating across more than a handful of different industries.
These include businesses like Marmon (diversified manufacturing), McLane (food distribution), Lubrizol (specialty chemicals), Shaw Industries (carpeting/flooring), Benjamin Moore (paint), Fruit of the Loom (apparel), Dairy Queen (restaurant), and See's Candies (food retail). This segment tends to grow through a combination of organic growth and acquisitions, generating a 17% revenue CAGR, with operating margins of around 8% on average, during the last decade. The business has traditionally required less than $2 billion in capital expenditures annually and carried an appropriate level of debt on its books, leaving it with a WACC of around 8%.
In our base-case scenario for Berkshire's manufacturing, service and retailing operations, we assume that the segment generates average annual revenue growth of around 5% longer term, based on a combination of organic growth and acquisitions. While this may seem conservative when compared with the historical results for the group, the five-year CAGR for revenue has been closer to 7%. We also feel that most of the businesses in the segment are driven by economic/population growth, with any additional growth coming from acquisitions (like the Lubrizol deal in 2011). With regard to profitability, the segment likely generated operating margins in the 7%-8% range overall this year, but given the nature of these businesses we feel it prudent to maintain margins at around 7% longer term. We also assume that capital expenditures increase as the size of this group overall increases. This leads to a $32,900 per Class A share (or $22 per Class B Share) value for the manufacturing, service and retailing operations.
Using an earnings multiple to triangulate and rationalize our fair value estimate for the company's manufacturing, service and retailing operations is a bit more tenuous. For starters, our four comparison companies-- Ingersoll-Rand (IR), ITT (ITT), Ashland (ASH), and CoreMark (CORE)--may be reflective of some of the bigger contributors--namely, Marmom, Lubrizol, and McLane--to earnings for the segment, but they ignore a whole host of companies and industries that may trade at higher or lower multiples than we are seeing with just these four comparable firms. But our implied earnings multiple is not widely divergent from the peer companies.
Finance & Financial Products Valuation
The finance and financial products segment is Berkshire's smallest segment, generating around 5% of annual operating income on average over the last decade. It includes two rental companies, XTRA (truck trailers) and CORT Business Services (furniture), and Clayton Homes, the leading producer and financer of manufactured homes in the United States, along with a collection of other financial assets. This segment has seen its ups and downs over the last decade, with the financial crisis and the great recession impacting revenue and profitability during the last five years. With equity and credit markets much more stable these days, and the economy gradually working its way toward a sustained recovery, we see the potential for 5% annual top-line growth in this segment longer term, with operating margins increasing from 17%-18% this past year to 19% over time.
All of which generates a $5,000 per Class A share (or $3 per Class B Share) value for the finance and financial products segment.
Our comparison companies for the finance and financial products segment should be a bit more conducive to comparisons based on book value per share. That said, with the segment having a negative book value the last couple of years, it is difficult to compare our fair value estimate for these operations with the peer group--CIT Group (CIT), Marlin Business Services (MRLN), and Rent-A-Center (RCII)--represented in our comparison table.
The implied earnings/fair value multiple of 14.1 times for the segment, though, does line up fairly well with the average price/earnings multiples (based on consensus estimates) for the three companies in our comparison table.
Berkshire's shares remain slightly undervalued in a market that appears to be fairly valued*
Our fair value estimate is equivalent to $187,500 per Class A share (or $125 per Class B share), reflective of a price to fair value multiple of around 0.85 times (inferring a more than 15% gain from today's trading prices). While not as large of a margin of safety as we would normally like to see in a firm with a medium uncertainty rating, we do note that Berkshire has effectively created a floor on the company's stock price by announcing that it would buy back both Class A and Class B shares at prices up to 120% of reported book value, which stood at $111,718 per Class A share (and $74 per Class B share) at the end of the third quarter of 2012. Furthermore, we anticipate that Berkshire's book value per share increased to at least $115,000 per Class A share (and $77 per Class B share) at the end of last year, meaning that Buffett would be willing to step in and buy the company's common stock at prices up to $138,000 per Class A share (and $92 per Class B share).
We do not anticipate that the firm's recently announced purchase of H.J. Heinz (HNZ) will have a material impact on our fair value estimate. Heinz is a classic Buffett firm that benefits from a significant competitive advantage caused by its strong brand presence. While the price is a little higher than we would have liked to see the firm pay, Berkshire is converting $12 billion in cash that was earning next to zero returns into an equity and preferred stake in the firm, the latter of which yields 9%.
That said, we would be remiss if we did not mention the fact that there are a few overhangs and headwinds that could hold down future stock appreciation--such as the ever increasing size of the firm and the eventual question of Buffett's succession. As such, we believe that a little caution is appropriate when considering Berkshire for any investment portfolio.
*Since this piece was authored, we raised our fair value estimate for Berkshire to $187,500 per Class A Share from $175,000 per share. The increase was due to slightly more optimistic assumptions about the firm's revenue, profitability, and cash flows.
Greggory Warren does not own (actual or beneficial) shares in any of the securities mentioned above. Find out about Morningstar’s editorial policies.