Arbitrage & Work-Outs

Concept Analysis

Definition & Origins

A work-out is an investment whose profit depends on a specific, publicly announced corporate event — a merger, sell-out, liquidation, reorganization, spin-off, or tender — rather than on a general rise in the stock's price. Buffett first defined the term for his partners in the 1957 letter, and the 1961 letter formalized it as the partnership's second investment category: securities "with a timetable," where the investor can estimate, within reasonable error limits, when he will get how much, and what might upset the applecart.

The practice came out of the Graham tradition. In the 1988 shareholder letter Buffett recalled participating in one such operation at age 24, while working in New York for Graham-Newman Corp.: Rockwood & Co., a Brooklyn chocolate company, offered to repurchase its stock with cocoa beans — 80 pounds of beans per share — exploiting an arcane provision of the 1954 Tax Code. Buffett spent several weeks buying shares, selling beans, and exchanging stock certificates for warehouse receipts; his only expense, he wrote, was subway tokens.

The term itself predates Wall Street's modern usage. As the 1988 letter explains, "arbitrage" once meant only the simultaneous purchase and sale of the same security in two different markets; since World War I the definition has expanded to "risk arbitrage" — the pursuit of profits from an announced corporate event. Buffett's partnership-era "work-outs" and Berkshire's later "short-term arbitrage commitments" are the same discipline at different scales: the announced event, not the market's mood, is the source of profit.

Core Ideas

The risk is the deal, not the market. A work-out bought after announcement carries little exposure to the general level of stock prices. Its real risk is that the expected development fails to materialize — anti-trust action, stockholder disapproval, a withheld tax ruling, a financing glitch. The 1957 letter put the distinction plainly: even a full-scale bear market should not hurt the market value of the partnership's work-outs substantially.

Small gross profits, decent annual returns. The spread between the post-announcement price and the deal value is usually a few percent. Buffett's friend called it getting the last nickel after the other fellow has made the first ninety-five cents. But the holding period is short and the outcome predictable, so the annualized math works: partnership work-outs typically returned 10% to 20% per annum on an all-equity basis, before any borrowed money.

Work-outs stabilize the whole portfolio. Because returns accrue irrespective of the Dow, a heavy work-out allocation looks brilliant in a declining year and dull in a roaring one. In 1962 the partnership's work-outs returned 14.6% while its generals lost 1.0% — the category, along with controls, "saved the day." In 1964, a big plus year for the Dow, the same category was a drag. Buffett never allocated between categories on a market forecast; the mix was, he admitted, largely an accident of availability.

Concentration plus modest leverage. Rather than spreading thin across fifty deals like the large arbitrage houses, Buffett held five to fifteen work-outs at a time, concentrating in a few situations per year. Because the category offered high safety in both eventual results and intermediate market behavior, he was willing to borrow against it — capped at 25% of partnership net worth — while regarding borrowed money against generals as very dangerous.

Errors are asymmetric, so discipline is everything. When a work-out falls through, the price does not drift down — it collapses back to its pre-announcement level. A practitioner cannot afford many mistakes. Hence the rigid rules: act only on publicly announced events, never on rumors; appraise each killjoy risk in advance; and stay out of deals that make no economic sense, because arbitraging those comes too close to playing the greater-fool game.

Practical Application

Texas National Petroleum (1962). The 1963 letter's appendix walks through a complete work-out, and Buffett stressed it was deliberately ordinary — "bread-and-butter," not a triumph. TNP, a small oil and gas producer, was being sold to Union Oil of California. Buffett heard rumors early in 1962 and did nothing; he moved only after the general terms were announced in early April. He then bought all three classes of securities: $264,000 of the 6½% debentures, 64,035 shares of common (proxy-estimated proceeds of $7.42 per share), and 83,200 warrants — about 13% of the entire issue — which worked out to $3.92 against their $3.50 exercise price. Stockholder approval risk was nil because the controlling stockholders had negotiated the deal; the only snag was a tax ruling, delayed because the University of Southern California's charitable status complicated Union Oil's financing. The deal closed October 31, months behind the original August-September estimate, and paid a little more than projected. The bonds earned about 20% per annum; the stock and warrants about 22%.

Hedging with short sales. The 1962 letter records $340,000 of securities sold short, mostly in conjunction with one late-year work-out where competition was briefly absent. The short leg eliminated the general market risk and locked in a 10%-or-better gross profit for a few months' tie-up of money.

The four questions (1988). At Berkshire the same checklist governed much larger commitments, spelled out in the 1988 letter: how likely is the promised event; how long will money be tied up; what chance is there of something better, like a competing bid; and what happens if the event fails. Arcata Corp. tested every one of them. Berkshire bought 400,000 shares around $33.50 after KKR agreed to pay $37.00 plus two-thirds of any recovery on a disputed redwood claim. The financing wobbled, KKR cut its bid, a rival group won at $37.50 plus half the claim — and Berkshire's 15% annualized return on the deal itself turned out to be the appetizer. The redwood claim, which Buffett had appraised at "somewhere between zero and a whole lot," settled in 1988 for $519 million, delivering an extra $29.48 per share.

Cash management, not career. In the Berkshire era arbitrage became an alternative to holding Treasury bills when cash exceeded good long-term ideas — a use Buffett valued partly because it removed the temptation to relax his standards for permanent investments. He sat out 1989 almost entirely because the deals on offer made no economic sense.

Common Misconceptions

Misconception 1: Work-outs are rumor or insider trading. Buffett's rule was the opposite: no action at the rumor stage, ever. He passed up the extra profit available in TNP before announcement because "that's somebody else's business, not mine." The partnership traded only on what it could read in the paper; at Berkshire, the 1988 letter repeated that they participate only in publicly announced transactions and do not guess takeover candidates.

Misconception 2: Work-outs are risk-free. They are market-independent, not risk-independent. Anti-trust action, stockholder disapproval, and withheld tax rulings are real killjoys, and when a deal breaks the market-value shrink is substantial. Buffett fully expected occasional losses and priced the category accordingly.

Misconception 3: Work-out returns are market-neutral in every sense. The category dampens relative performance in both directions. In 1964, with the Dow up strongly, work-outs were an explicit drag on the partnership's results; the 14.6% they returned in the 1962 decline was the mirror image. An investor who cannot tolerate lagging a bull market cannot run a heavy work-out book.

Misconception 4: The annualized percentages scale. They do not. As the partnership grew, good work-outs large enough to matter grew scarce. The 1965 letter called the business "very spasmodic" — $6 million average capital employed, over half the year's profit from a single situation — and the 1966 letter doubted the old 30-40% allocation would ever return. Size eventually pushed Buffett toward controls and, later, wholly-owned businesses.

Thought Evolution

Related Concepts

Case Companies