Charlie Munger on Capital Allocation

At the 2023 DJCO meeting, Munger was asked about the bank stocks in the Daily Journal securities portfolio. The question was pointed: Berkshire had unloaded its bank stocks, and if those positions were not good enough for Berkshire shareholders, why were they good enough for Daily Journal shareholders? Munger's answer was structural. He might have different ideas than Berkshire, he said. If you owned marketable securities within a corporation located in California, you paid huge state and federal taxes if you sold things at a big gain, and that affected the willingness to sell. He then made a striking disclosure: those bank stocks he had bought on the bottom tick in the foreclosure crisis. Literally, Munger said, it was the bottom tick. They were practically all gain now, so he would immediately give the government forty-some percent of everything he sold out of those bank stocks. They were producing dividends that were almost tax free. Based on what he would get if he sold them and the return he was getting out of the dividends, he said, it's not so bad for us. The answer was that Daily Journal was not in a normal position. All factors considered, they were willing to hold them for a while. The decision, in other words, was not a vote against the underlying thesis. It was a vote for tax discipline. The big disadvantage in having a huge layer of federal corporate taxes and state taxes between the company and any money it made - in a state like California, especially - was that it trapped gains inside the corporate shell. Munger was telling the room that tax friction is a real input into hold-versus-sell decisions, that the bottom-tick buy had produced a position where the after-tax math of selling was inferior to the after-tax math of holding, and that the rational investor factors that into the decision rather than pretending it doesn't exist.

At the 2019 Berkshire annual meeting, Munger was asked about share repurchases. The question probed the precision of Buffett's buyback threshold and whether Berkshire would be more liberal in repurchasing its own stock. Munger's answer was deliberately imprecise. He told the audience that he was a little more liberal in repurchasing shares than Buffett, and that the question of un-precision in railroading was a related problem - at some point, in a complicated operation, you accepted that you were operating with judgment rather than measurement. The point Munger was making was that capital allocation at Berkshire scale was not a marks-to-the-penny exercise. Repurchasing shares below intrinsic value was a clear duty when the price was clearly below the estimate; the difficulty was that intrinsic value itself was an estimate, not a quote. He told the room that pretending to more precision than the business actually allowed was itself a form of misjudgment. The honest framing was that Buffett and Munger had a range for intrinsic value, and they repurchased aggressively when the market price fell well below the low end of that range. The corollary was a critique of the modern buyback fashion. Munger noted that, historically, companies had refused to buy back their stock when it was a very good idea and were buying it back aggressively when the stock was so high that doing so was frequently a bad idea. He welcomed the audience to adult life - this is the way it is. The observation was that corporate buyback behavior was pro-cyclical, driven by the same incentive biases that drove every other form of capital allocation. The disciplined operator did the opposite: he bought back stock when the price was low and refrained when the price was high, regardless of what the Street was telling him about the optics.

At the 2015 Daily Journal meeting, Munger described the company's pivot from print legal journalism to court-automation software as the equivalent of trying to climb Half Dome in Yosemite with one arm and one leg. The franchise had been a wonderful business - a monopoly on prompt appellate court decisions, year after year of price increases. Then the internet came along and destroyed the position. Daily Journal's circulation went way down, and the publishing business shrank with it. The decision to replace the dying newspaper with software sold to courts and government agencies was, Munger said, probably not a terribly good decision at the time. They tried it anyway. A great boom in foreclosure notices temporarily flooded Daily Journal with revenue, and the company used that transient cash to build the software business partly by purchase and partly by self-development. The odds were against them, Munger admitted. He used the rock-climbing term five-eleven to describe what they were attempting - a route that is not really possible, but that occasionally somebody does climb. He told shareholders that for some strange reason Daily Journal was now about halfway up Half Dome with its one arm and one leg. Software revenues had crossed the level of the traditional business. He was candid that the cost had been heavy and would continue to be heavy, but said he thought about the spend the way Jeff Bezos does: there is no point in being rich if you don't use it to compete effectively. He closed by saying that the kind of business they were building was so hard that competitors like Microsoft hated it. That difficulty, in Munger's calculus, was the only reason the opportunity existed at all.

In the 2007 Berkshire shareholder letter, Buffett - crediting Munger throughout - used the See's Candies acquisition as the textbook case for what a brand franchise actually does to a business. Berkshire had bought See's in 1972 for $25 million, against an asset value of about $8 million and pre-tax earnings of about $4 million. The price looked full to the traditional cigar-butt investor, and Buffett had initially hesitated. Munger had pushed him to pay it, arguing that the franchise was worth the premium because the brand could raise prices year after year without losing volume. The retrospective made the math visible. See's had generated pre-tax earnings cumulatively in the many hundreds of millions of dollars in the years since purchase, on the original $25 million base. The asset base had grown only modestly. The incremental capital required to grow the business had been tiny relative to the cash thrown off. The whole return had come from the brand's pricing power, not from reinvestment. That, Munger and Buffett were saying, is what a real moat looks like - the cash grows faster than the asset base because customers keep paying up for the name. Munger's investment lesson, distilled in the 2007 letter, was that See's taught Berkshire to look past the cigar-butt habit and toward the great franchise. The intangibles - brand, distribution, customer loyalty, pricing power - were not a speculative add-on to intrinsic value. They were the source of it. The companies that grew cash faster than assets were the companies that compounded intrinsic value per share, and the only way to find them was to look at the qualitative strengths that traditional accounting did not capture. See's was the school. Every later Berkshire acquisition - Coca-Cola, Gillette, GEICO in full - was a graduate of that school.

It merely resulted in a reclassiÑcation from unrealized gains to retained earnings, another component of shareholders' equity. This supplementary breakdown of earnings diÅers somewhat from that used in audited Ñnancial statements which follow standard accounting convention.foregoing

in income taxes payable, on its consolidated balance sheet. Thus, the entire after-tax gain on the non-cash merger had been reÖected in the unrealized gain component of Wesco's shareholders' equity as of September 30, 2005. That amount was merely switched from unrealized gain to retained earnings, another component of shareholders' equity. This accounting entry had no economic eÅect on Wesco, and you should ignore it when you are evaluating Wesco's 2005 earnings. Consolidated Balance Sheet and Related Discussion As indicated in the accompanying Ñnancial statements, Wesco's net worth, as accountants compute it under their conventions, increased to $2.40 billion ($337 per Wesco share) at yearend 2006 from $2.23 billion ($313 per Wesco share) at yearend 2005. The main causes of the increase were appreciation in fair value of investments, and net operating income after deduction of dividends paid to shareholders. The foregoing $337-per-share book value approximates liquidation value assuming that all Wesco's non-security assets would liquidate, after taxes, at book value. Of course, so long as Wesco does not liquidate, and does not sell any appreciated securities, including the PG shares Wesco received in connection with PG's acquisition of Gillette in 2005, discussed above in the section, ""Realized Investment Gains,'' Wesco has, in eÅect, an interest-free ""loan'' from the government equal to its deferred income taxes, subtracted in determining its net worth.

It merely resulted in a reclassiÑcation from unrealized gains to retained earnings, another component of shareholders' equity. This supplementary breakdown of earnings diÅers somewhat from that used in audited Ñnancial statements which follow standard accounting convention. The foregoing supplementary breakdown is furnished because it is considered useful to shareholders.

Accounting standards promulgated by the Financial Accounting Standards Board require that the fair (market) value of shares received in such an exchange be recorded as the new cost basis as of the date of the exchange, with the difference between the new basis and the historical cost realized in the audited financial statements as an investment gain. For tax return purposes the exchange is recorded at the original cost of the securities exchanged; no gain is reported, and no taxes are yet due. Although the realized gain had a material impact on Wesco's reported earnings, it had no impact on Wesco's shareholders' equity. Wesco carries its investments at fair value, with unrealized appreciation, after income tax eÅect, included as a separate component of shareholders' equity, and related taxes included in income taxes payable, on its consolidated balance sheet. Thus, the entire after-tax gain on the non- cash merger had been reÖected in the unrealized gain component of Wesco's shareholders' equity as of September 30, 2005. That amount was merely switched from unrealized gain to retained earnings, another component of shareholders' equity. This accounting entry had no economic eÅect on Wesco, and you should ignore it when you are evaluating Wesco's 2005 earnings. Consolidated Balance Sheet and Related Discussion As indicated in the accompanying Ñnancial statements, Wesco's net worth, as accountants compute it under their conventions, increased to $2.

At the 2003 Wesco annual meeting - held in the same week as the now-famous Berkshire derivatives warning - Munger reflected on what made Berkshire's structure durable. He told the audience that the Berkshire conglomerate model worked because it had almost no corporate bureaucracy. There was practically nobody at headquarters. The people running the operating businesses were sensible people who were left alone to run them. The absence of bureaucracy was, in Munger's view, a huge advantage - not a management fad but a structural property of the conglomerate that compounded year after year. He paired the structural observation with a warning. Bureaucracy, Munger said, breeds failure and stupidity. How could it be otherwise? The point was that bureaucratic organizations systematically destroyed the judgment of the people inside them, because the bureaucratic structure rewarded process over outcome, compliance over insight, and risk-avoidance over capital allocation. A company that allowed a bureaucracy to grow was, in Munger's framing, slowly converting its smartest operators into the kind of people who could not make a real decision even when one was needed. The takeaway for the room was that capital allocation discipline and organizational discipline were the same problem. Berkshire's edge was not just that Buffett and Munger had good judgment; it was that the structure they had built protected the judgment of the operators below them from being bureaucratized away. The same lesson, Munger implied, applied to the smaller Wesco conglomerate - KBS, Wes-FIC, Precision Steel - and to any organization that wanted to compound intrinsic value over decades. Keep the headquarters empty, keep the operators in charge, and refuse to grow the corporate center in the name of oversight.

Those shares, carried on Wesco's balance sheet at yearend 1999 at a market value of $1.4 billion, were sold in 2000, giving rise to the principal portion of the $852.4 million of after-tax securities gains realized by Wesco in 2000, versus no gains or losses realized in 2001. Although the realized gain had a material impact on Wesco's reported earnings for 2000, it had a very minor impact on Wesco's shareholders' equity. Inasmuch as the greater portion of the realized gain had previously been reÖected in the unrealized gain component of Wesco's shareholders' equity, the amount was merely switched from unrealized gains to retained earnings, another component of shareholders' equity. Consolidated Balance Sheet and Related Discussion As indicated in the accompanying Ñnancial statements, Wesco's net worth, as accountants compute it under their conventions, decreased to $1.91 billion ($269 per Wesco share) at yearend 2001 from $1.98 billion ($278 per Wesco share) at yearend 2000. The foregoing $269-per-share book value approximates liquidation value assum- ing that all Wesco's non-security assets would liquidate, after taxes, at book value. Perhaps this assumption is too conservative.liquidation

when Freddie Mac shares could be lawfully owned only by a savings and loan association. Those shares, carried on Wesco's balance sheet at yearend 1999 at a market value of $1.4 billion, were sold in 2000, giving rise to the principal portion of the $852.4 million of after-tax securities gains realized by Wesco in 2000, versus $7.3 million, after taxes, realized in 1999. Although the realized gains materially impacted Wesco's reported earnings for each year, they had a very minor impact on Wesco's shareholders' equity. Inasmuch as the greater portion of each year's realized gains had previously been reÖected in the unrealized gain component of Wesco's shareholders' equity, those amounts were merely switched from unrealized gains to retained earnings, another component of shareholders' equity. Consolidated Balance Sheet and Related Discussion As indicated in the accompanying Ñnancial statements, Wesco's net worth, as accountants compute it under their conventions, increased to $1.98 billion ($278 per Wesco share) at yearend 2000 from $1.90 billion ($266 per Wesco share) at yearend 1999. The foregoing $278-per-share book value approximates liquidation value assum- ing that all Wesco's non-security assets would liquidate, after taxes, at book value. Probably, this assumption is too conservative.

The loss widened in 1999 because fewer dividends were received during the year after forced conversion of preferred stock of Citigroup Inc. (""Citigroup'') into lower-dividend-paying common stock. The ""other 'normal' net operating income or loss'' Ñgures for 1999 and 1998 also include intercompany charges for interest expense ($353,000 and $102,000 after taxes, respectively) on borrowings from Wes-FIC. This intercompany interest expense does not aÅect Wesco's consolidated net income inasmuch as the same amount is included as interest income in Wes-FIC's ""normal'' net operating income. ""Other 'normal' net operating income or loss'' beneÑted in 1999 by about $800,000 caused by reversals of reserves for possible losses on sales of loans and tag-end real estate, expensed in prior years. Net Securities Gains and Losses Wesco's earnings contained securities gains of $7,271,000, after income taxes, for 1999, versus $33,609,000, after taxes, for 1998. Although the realized gains materially impacted Wesco's reported earnings for each year, they had a very minor impact on Wesco's shareholders' equity. Inasmuch as the greater portion of each year's realized gains had previously been reÖected in the unrealized gain component of Wesco's shareholders' equity, those amounts were merely switched from unrealized gains to retained earnings, another component of shareholders' equity.

Inasmuch as the greater portion of each year's realized gains had previously been reÖected in the unrealized gain component of Wesco's shareholders' equity, those amounts were merely switched from unrealized gains to retained earnings, another component of shareholders' equity.

Daily Journal Corporation's late-1990s filings are notable for what they do and do not show. The company had, under Munger's chairmanship, avoided the speculative derivatives exposure that had destroyed several of its peers in the savings-and-loan and publishing-adjacent industries during the savings-and-loan crisis. The 10-K filings disclose a conservatively-financed publisher with a real moat - the appellate-decisions monopoly - and no exposure to the structured products that had ruined other ostensibly conservative companies in the same region. Munger's role at DJCO throughout the 1990s was, in effect, the same role he played at Berkshire: the disciplined refuser. He had refused to let Daily Journal take on the leverage that the cheap-money years of the mid-1990s had tempted other small public companies to take on. He had refused to chase the incremental yield that derivatives contracts appeared to offer. He had insisted that the company finance itself conservatively, hold its franchise honestly, and let the cash earnings of the legal publishing monopoly compound rather than leveraging them up in the name of growth. The retrospective lesson, visible in the 1997 10-K, was that avoidance was the operating decision. The companies that failed in the savings-and-loan crisis had not failed because they were stupid; they had failed because they had taken on exposure they did not need to take on, in pursuit of returns they did not need to pursue. Daily Journal, under Munger, had refused the exposure and survived the crisis with its franchise intact and its balance sheet clean. The same discipline would, two decades later, allow Daily Journal to pivot into court-automation software with the financial strength to absorb the long, slow, expensive slog of building that business. The avoidance had bought the optionality.

At the 1997 Berkshire annual meeting, Buffett and Munger were asked to explain how they decided whether to buy a business. The afternoon session produced one of the clearest distillations of the joint filter the two men applied. The first cut was whether they could genuinely understand the business - not the stock, not the industry narrative, but the unit economics, the moat, and the trajectory of intrinsic value over a decade. If that test was failed, they moved on without remorse. Munger's contribution to the answer was characteristically about the negative space. The discipline was less about saying yes to the right business and more about saying no to the wrong ones, fast. He told the audience that they did not have a list of businesses they wanted to be in; they had a list of businesses they refused to be in. The screening-by-exclusion was the actual operating system. Most deals, in most years, did not pass the first screen. The second cut was management. They had to be people Buffett and Munger would trust to run the business correctly without supervision - not people they would have to monitor, but people they could empower and leave alone. Munger's standard for managers was character first, judgment second, energy third. He told the room that you could not fix a character problem with compensation, and you could not fix a judgment problem with effort. The filter was severe, which was the whole point of having a filter at all.

Accounting standards promulgated by the Financial Accounting Standards Board require that the fair (market) value of shares received in such an exchange be recorded as the new cost basis as of the date of the exchange, with the diÅerence, after appropriate reserves for future income tax on the gain, recognized in the Ñnancial statements as a realized after-tax gain. For income tax purposes the exchange is recorded at the original cost of the securities ex- changed; no gain is reported on the tax return, and no taxes are yet due. Although the realized gain had a material impact on Wesco's reported earnings, it had a very minor impact on Wesco's shareholders' equity. Inasmuch as $48,504,000 of the after-tax gain had previously been reÖected in the unrealized gain component of Wesco's shareholders' equity as of September 30, 1997, that amount was merely switched from unrealized gains to retained earnings, another component of share- holders' equity. Convertible Preferred Stockholdings At the end of 1997, Wesco and its subsidiaries owned $52 million, at original cost, in convertible preferred stocks of Travelers Group Inc. (""Travelers'') and US Airways Group, Inc. (""US Air''). The Travelers preferred stock was received in late 1997 (see the preceding section) in exchange for the Wesco group's remaining shares of Salomon Inc preferred stock, which originally cost $40 million, and whose cost was adjusted upwards to $90 million as of the date of the exchange.

At Harvard in 1995, Munger opened his talk on human misjudgment with what he considered the most underappreciated driver of bad decisions in the entire literature: incentive-caused bias. He told a doctor story from his own youth in Lincoln, Nebraska. A doctor there had been sending bushel baskets of normal gallbladders down to the pathology lab at the leading hospital, and the quality-control machinery of community medicine had taken about five years longer than it should have to remove him from the staff. Munger asked an older doctor who had participated in the removal whether the man had consciously been running a maiming-and-murder-for-profit scheme. Hell no, came the answer - the man had convinced himself the gallbladder was the source of all medical evil, and that if you loved your patients you couldn't get it out fast enough. Munger's point was that incentive bias operates with full force even in people you would gladly marry into your family. It is present in every profession and in every human being. He pushed the room to generalize from the example: sales presentations and brokers of commercial real estate, in his experience, were never even within hailing distance of objective truth. The same mechanism that produced the gallbladder surgeon produces the mispriced collateralized product, the pumped-up sell report, and the cost-plus contract that rewards running the budget up rather than down. He closed the loop with the cash register story. Patterson's little store was being stolen blind, the cash register fixed it, profit appeared instantly - and Patterson then closed the store and went into the cash register business. Munger's conclusion: people who invent things like cash registers, which make most bad behavior hard, are some of the effective saints of our civilization. The cash register was a moral instrument when it was created. Designing systems that contain incentive bias is therefore one of the highest-leverage forms of ethical action a society can take.