Sunday, April 24, 2011

David Sokol and Berkshire Internal Controls

The David Sokol situation has raised any number of issues -- not the least of which is the nature and efficacy of Berkshire internal controls. Interestingly, Sokol discussed them in some detail during his March 31 CNBC interview.

JOE Kernan: There's a lot of, there's other employees as well. This brings up, or begs the question to a lot of people about what Berkshire's internal controls are on employee purchases —


SOKOL: I mean, Warren and (Berkshire CFO) Marc Hamburg furnish us a list of stocks that we are restricted on. Any companies that, you know, that apparently Warren or historically Lou Simpson, or now Todd Combs, or whoever, is invested in, ah, that, you know, we can't ever buy or sell without first contacting Mark. This certainly wasn't one of those companies.....


SOKOL: That's right. In fact, I have no authority whatsoever. I couldn't spend a dollar of Berkshire's money buying, buying a security tomorrow.

It is not publicly known if Berkshire audits the trading activity of its named insiders with respect to companies on the 'list'. But if not, it could be accomplished in a relatively short time. It is reasonable to believe it could be done in a matter of hours -- with the cooperation of insiders, if it hasn't already been done.

Piecing together Berkshire's investment process, it is based on extreme concentration of decision making regarding capital allocation. Buffett (and Munger) make the decisions, with Simpson and Combs running a small portfolio.

Buffett is also famous for what is an unusual decision making style. These are well known and include:

1. Doesn't do auctions or hostile takeovers.
2. Doesn't talk (extensively) without a selling price.
3. Isn't interested in pitches involving publicly traded companies.
4. Extensively uses publicly available financials.
5. Uninterested in the types of work done by most stock analysts.

Of special note is this comment which has been included in his published acquisition criteria for years:
We are not interested, however, in receiving suggestions about purchases we might make in the general stock market.

Considering this as well as the new information that Berkshire uses a control system for public companies based on a list -- the effectiveness of Berkshire's systems in the past are explainable.

If Berkshire insiders refrain from trading in stocks on the list, and given that Buffett/Simpson/Combs are the only decision makers regarding publicly held companies, then the only real possible candidates for "front running" are Berkshire acquisitions of public companies in which Berkshire currently has no position.

Over the last decade, that has included Clayton Homes and Burlington Northern (which would have presumably been on the 'Hamburg' list) which makes for a very short list.

This approach is an interesting variant on the well known separation of duties principle. It is reasonable to assume that Berkshire has established industry standard accounting controls. Their overall thrust is to eliminate the opportunities for insider trading.

Contrast this with a typical asset manager. They have their own analysts, use proprietary data, methods, and processes, and delegate significant parts of the process over a number of employees.

Berkshire's home office has only 20 employees, none of them (except Marc Hamburg) senior executives. I would assume that they are aware of numerous proposals -- so many that it is likely that they would be of little value to typical traders. Among other things, Berkshire is frequently mentioned in the press as a potential savior of a doomed company -- the last hope of imploding companies.

So -- how does this relate to Sokol?

One thought is that Sokol, who has no experience in financial or investment firms outside the totally unique environment of Berkshire, inferred that anything not proscribed by the Hamburg list was fair game. Of course this is speculative, but Sokol's behavior is very difficult for me to explain any other way.

More specifically, he lacks the usual motivation and associated behavior associated with insider trading. He doesn't need the money, and the amounts involved, while significant, are not material to Sokol's personal wealth -- and certainly not sufficient for him to take huge personal risks. More importantly, he made no effort to hide his trading.

His actions are consistent with that of a person that was simply unaware of the significance of his trading. This is not an excuse, but a hypothesis.

The 'Hamburg' process -- a proscribed list of publicly traded securities -- may be a clue to how an intelligent man could behave so recklessly.

Berkshire has a system that explicitly proscribes insiders from trading in specific securities. The also have a principle based standard that no one should do anything that they would be uncomfortable reading about on the front page of the New York Times. Sokol has now had the opportunity to to do exactly that. Whatever else is said, it couldn't have been comfortable.

Saturday, February 26, 2011

Normalized Earnings - Berkshire Hathaway 2010 Annual Report

Warren Buffett is fantastic at making arcane accounting accessible to an average investor. But it can be hard to go from his big picture analogies to the detailed figures that make up the published financials.

This year he introduced an important new metric to the report -- Normalized Earnings. He offers no details on how he came up with the figure, ($12 billion), but it is simply another approach to separate the noise of short term fluctuations from the underlying, core value of the firm.

This figure is another approach to complement the two principle metrics he has always used. That is -- for the short term, operating earnings and for the long term, change in book value (shareholder equity).

Buffett defines normalized earnings as, "... a year free of a mega-catastrophe in insurance and possessing a general business climate somewhat better than that of 2010 but weaker than that of 2005 or 2006." And he has selected a figure of $12 million as his estimate.

He once again discusses the long term (46 years) metric he has consistently shown at the beginning of each annual report -- change in Book Value.

"To eliminate subjectivity, we therefore use an understated proxy for intrinsic-value – book value – when measuring our performance....Yearly figures, it should be noted, are neither to be ignored nor viewed as all-important. The pace of the earth’s movement around the sun is not synchronized with the time required for either investment ideas or operating decisions to bear fruit."

To demonstrate the value of this metric over intermediate time periods, the report (page 5) includes the exact same data aggregated into rolling 5 year periods. This exhibit is striking in regard to the extent to which the figures stabilize and present a much clearer picture of past performance.

This works particularly well for Berkshire, since it doesn't pay dividends. Book value is the aggregate of the entire financial history of the firms -- and includes all capital gains -- both realized and unrealized. And all the exceptions that are typically excluded from operating earnings.

Later in the report, a discussion of operating earning vs net earnings repeats numerous prior discussions:

"Operating earnings, despite having some shortcomings, are in general a reasonable guide as to how our businesses are doing. Ignore our net income figure, however. Regulations require that we report it to you. But if you find reporters focusing on it, that will speak more to their performance than ours."

Net earnings are a GAAP figure that is meaningful for most businesses, but is severely flawed for a firm with a large investment portfolio and atypical derivative contracts. Operating earnings exclude the change in derivative liabilities as well as realized capital gains and losses.

Since 2006, Berkshire has issued a press release with the non GAAP operating figures reconciled with the GAAP net income figures. They have included this statement (from 2007):

"In our earnings summary, we distinguish between what we call “operating earnings” and investment and derivative gains/losses. Berkshire possesses a huge reservoir (about $35.5 billion on June 30, 2007) of pre-tax unrealized investment gains. The cashing of these in any given quarter (or the realization of losses, for that matter) can materially distort net income figures as well as comparisons between periods. We do not wish investors to mistakenly focus on a bottom-line number affected by large investment gains that do not stem from economic accomplishments during the reporting period and that have no concurrent impact on the intrinsic value of the company. Both trends in our operating businesses and their health are best judged by income before all investment gains or losses."

The only thing missing from this explanation is the implicit assumption that the only additional information is the latest quarter's data. Everything else has already been published and presumably included in valuation of the company. Another way of stating this is that over a single quarter, operating earnings are much more appropriate for a firm like Berkshire than GAAP net income. The quarterly press releases stress this both in quarters where the headline net income figure is higher as well as lower than operating earnings.

Page 33 of this years annual report:

This is a GAAP exhibit and reconciles the balance sheet to the income statement -- and illustrates the various items that impact the change in book value other than net income. The largest is 'Other Comprehensive Income" and includes unrealized capital gains -- which is shown in detail on page 33:

Note that over the 3 year period, the largest items - unrealized capital gains/losses and their tax impact - net to a modest figure. This reflects the market crash in 2008 as well as the stock market recovery in 2009 and 2010.

Net income is clearly a better proxy for earning power (for lack of a better term) than comprehensive income. However, it includes realized capital gains/losses and derivative gains and losses. And also remember that GAAP is designed for all businesses in all industries -- and most businesses don't have any derivatives and have modest capital gains/losses.

" After-tax investment and derivative gains/losses were $1.87 billion in 2010, $486 million in 2009, $(4.65) billion in 2008, $3.58 billion in 2007 and $1.71 billion in 2006."(page 27)

The net income for 2010, 2009, and 2008 of $13.0, $8.1, and $5.0 billion, less derivative and investment gains/losses, are the operating earnings of $11.1, $7.6, and $9.6 billion. The three year average net income was $8.7 billion vs average operating earnings of $9.4 billion.

The 2010 operating earnings of $11.1 billion as well as the three year average of $9.4 billion provide a lot of support for the new metric, normalized earnings, estimated at $12 billion.

[end of part 1]