The Outsiders

1. 1Replace size metrics with one scorecard: compound annual return to shareholders over your tenure (or the period you are judging), the same figure for peer companies, and the same figure for the S&P 500. These are the three numbers the source says you need to judge greatness. Revenue, headcount, and earnings growth are secondary.
2. 2Set the context before reading the result. Note the start and end dates and what the market did in between. Welch's 20.9% a year ran through a bull market in which the S&P averaged about 14%, roughly double its long-term average. A 20% return in that window is not a 20% return in a window containing several bear markets.
3. 3Benchmark against a broad peer set facing the same conditions, not one favorite rival. Peers are dealt similar hands, as in duplicate bridge, so the gap between you and them reflects management decisions rather than external forces. For a commodity business, compare against others operating at the same commodity price.
4. 4Treat capital allocation as the main job and rank every use of cash against the others: capex, acquisitions, dividends, debt paydown, and repurchases of your own stock. The test for each is the expected increase in per-share value, not the growth it adds to the company.
5. 5Measure per-share value on cash flow, not reported earnings. Judge each option by what it adds to long-run cash flow per share.
6. 6Run a standing price-versus-intrinsic-value comparison on your own stock, as Chabraja did at General Dynamics. When the price is meaningfully below your estimate of value, repurchase aggressively. Singleton always bought every share tendered because he had done the analysis and held conviction. When the price is high, issuing stock becomes an option.
7. 7Size buybacks to the opportunity. Singleton's tenders ranged from 4% to 66% of book value, and Anders retired 30% of General Dynamics' shares in one 1992 tender. Do not default to token buybacks if the discount is large.
8. 8Be willing to wait. If no project clears the per-share hurdle, hold cash. Dick Smith waited about a decade at General Cinema. Stay active when others are fearful: every outsider CEO ran a large buyback or a series of large acquisitions during 1974–1982.
9. 9Keep overhead from eroding per-share value. Singleton ran a wafer-thin corporate staff with authority concentrated in unit general managers, and he skipped courting analysts and journalists to protect his time for allocation decisions.

Henry Singleton at Teledyne is the source's model case. From 1963 to 1990 he delivered a 20.4% compound annual return. A dollar invested in 1963 grew to $180, against $27 for a broad group of conglomerates and $15 for the S&P 500, which is over twelve times the index. His method was allocation, not size. Earlier, while building Teledyne, he issued stock at an average P/E above 25. In the 1970s he reversed course and bought back stock at an average multiple under 8, spending $2.5 billion in tender offers of 4% to 66% of book value. The tenders were mostly oversubscribed, and he bought all shares offered. The buybacks earned a 42% compound annual return for continuing shareholders. From 1971 to 1984, revenues and net income kept growing while the share count shrank, producing a fortyfold increase in earnings per share. Eventually he bought in over 90% of the shares. He also avoided dividends, emphasized cash flow over reported earnings, and never split the stock.

• If your stock trades well below your conservative estimate of intrinsic value, then repurchase heavily, sized to the discount, and take all shares offered in a tender.
• If your stock trades at a high multiple relative to value (Singleton issued above 25x earnings), then issuing stock for acquisitions is a reasonable source of capital. Do not buy back at that price.
• If no acquisition or project clears the per-share return hurdle, then wait and hold cash rather than buy size. Dick Smith waited a decade.
• If a return looks impressive, then check it against peers and the S&P over the same dates before calling it good. Absolute returns during a bull market prove little.
• If an option raises revenue or reported earnings but not cash flow per share, then reject it or rank it below options that raise cash flow per share.

• Benchmarking on growth: celebrating sales, profit, or headcount increases. The tell is that your scorecard has no per-share or relative-return line.
• Ignoring the market's context: crediting yourself for returns earned in a bull market. The tell is that you can't state the S&P return over the same period.
• Cherry-picked peers: comparing against weak companies or a narrow set. The tell is that your peer group changes whenever your relative result worsens.
• Buying back stock on a schedule or at any price, rather than against intrinsic value. The tell is that repurchases are steady regardless of the P/E, or stop in downturns when prices are lowest.
• Passivity in bad markets: sitting on your hands during fear, such as the 1974–1982 period when equities were declared dead. The tell is that your repurchase and acquisition activity dries up exactly when prices are most attractive.
• Letting Wall Street's views set strategy by courting analysts and the press. The tell is that a major decision is shaped more by anticipated commentary than by per-share math.

• ▸Measurement period (start and end dates): fix the window and note the market conditions in it (bull, bear, mixed).
• ▸Company compound annual return to shareholders: enter the figure for that window.
• ▸Peer group compound annual return: list the broad set of peers facing the same conditions and enter their return.
• ▸S&P 500 compound annual return, same window: enter it, then state your multiple versus the index.
• ▸Cash flow per share, start vs. end: use cash flow, not reported earnings, and note any gap between the two.
• ▸Current price vs. conservative intrinsic value estimate: enter both and the discount or premium.
• ▸Candidate uses of capital ranked by expected per-share value added: list capex, acquisitions, dividends, buyback, and debt paydown with the expected effect on each.
• ▸Buyback or issuance decision and size: state the action, size as a share of book value or shares outstanding, and the P/E multiple you would pay or receive.
• ▸Wait condition: write the hurdle that must be cleared before deploying cash, and what you do if nothing clears it.

1. 1Fix the tenure window first. Use the CEO's actual start and end dates (Welch: 1981 to 2001; Singleton: 1963 to 1990, which the passages call both 'almost thirty years' and 'twenty-eight years'). Do not pick a convenient window.
2. 2Compute compound annual TSR to shareholders over that full window. Revenue, profit, and employee growth do not count; the passages treat per-share value as the ultimate barometer.
3. 3Run the same calculation over the identical dates for the S&P 500. Express the result as a multiple of the index's cumulative return per $1 invested (Singleton: $180 vs. $15, about 12x).
4. 4Build the peer benchmark from a broad universe of companies dealt similar hands, usually the same industry. For Singleton, that was a broad group of conglomerates ($27 per $1). Companies in the same industry face the same external forces, so the gap reflects managerial ability. Gold miners in 2000 and 2011 face different commodity prices and cannot be compared.
5. 5Require both tests to pass. A CEO must beat the market and the peers by a wide margin. Beating only one benchmark is not 'great' under this definition.
6. 6Adjust for context. Note the market regime over the window, such as Welch's 1982-2000 bull market, where the S&P averaged about 14%, roughly double its long-term average. A 20% return in a bull market ranks below the same return across several bear markets.
7. 7Check duration and sustainability. Longer tenures with the same relative outperformance are stronger evidence. Singleton's 28 years against Welch's 20 counts in his favor.
8. 8Report the result as three numbers: the CEO's compound annual return, the peer return, and the S&P return over the same window. The passages say these three are all you need.

Jack Welch ran GE from 1981 to 2001 and delivered a 20.9% compound annual return, so $1 became $48. That outperformed the S&P 500 by 3.3 times. The tenure coincided almost exactly with the bull market that began in late 1982 and ran largely uninterrupted to early 2000, when the S&P averaged about 14% a year. Henry Singleton ran Teledyne for almost thirty years and returned 20.4% a year. $1 invested in 1963 was worth $180 by 1990, against $27 in a broad group of conglomerates and $15 in the S&P 500, which is over twelve times the index. His peers averaged only 11% a year. Singleton's nominal return is slightly lower than Welch's. But his multiple of the market is far higher, his tenure is longer, and his window included several protracted bear markets. On the three-number test, Singleton is the greater CEO.

• If a CEO's return is high but the S&P 500 averaged double its long-term rate over the window, then discount it and judge the multiple of the market rather than the absolute rate.
• If a CEO beats the index but only matches or modestly beats peers, then do not call the record great. Both benchmarks must be cleared by a wide margin.
• If the peer set spans industries with different external drivers, such as commodity prices, then narrow it to companies facing the same conditions before comparing.
• If two CEOs have similar compound rates, then rank by the multiple over market and peers, then by tenure length and market difficulty.
• If the evidence is revenue, earnings, or magazine coverage rather than per-share return, then set it aside and recompute TSR.

• Judging by growth in revenues or profits. The tell: the case rests on size, sales growth, or headcount rather than per-share value.
• Equating fame with performance. The tell: your sample is the Fortune 100 or magazine-cover CEOs. Singleton never appeared on the cover of Fortune.
• Using the absolute return without market context. The tell: a 20% figure is cited with no S&P comparison, and the window falls in a bull market.
• Choosing a sloppy peer group. The tell: the peers are in unrelated industries or face different commodity or external conditions, so the gap is not due to managers.
• Cherry-picking start and end dates. The tell: the window does not match the CEO's actual tenure, or it starts at a trough or ends at a peak.

• ▸CEO and company: name the executive and the firm.
• ▸Tenure start and end dates: use the actual dates in office, not a flattering window.
• ▸Company compound annual TSR: calculate it over the full tenure.
• ▸$1 invested value: show what $1 at start is worth at end for the company.
• ▸Peer group and rationale: list the companies and explain why they face the same conditions.
• ▸Peer compound annual return and $1 value: calculate it over the identical dates.
• ▸S&P 500 compound annual return and $1 value: calculate it over the identical dates.
• ▸Multiples: company $1 value divided by S&P $1 value, and by peer $1 value.
• ▸Market context note: record whether the window was a bull market or included bear markets, and adjust your read.
• ▸Verdict: state whether the CEO clearly beat both benchmarks by a wide margin and sustained it over a long tenure.

1. 1Name the institutional imperative before you act. Buffett's 1986 Berkshire letter describes it as the corporate equivalent of teenage peer pressure, a force that pushes CEOs to imitate peers. For each major decision, write down what your peers are doing and whether that is driving your inclination.
2. 2Reduce the business to its core economics. The outsiders cut through peer and press chatter to the core economic characteristics of their businesses. Write a one-page statement of how the business produces cash. Strip out whatever debt levels, taxes, capex, or acquisition history do to reported numbers.
3. 3Make cash flow your scorecard, not quarterly reported net income. Net income is distorted by debt, taxes, capital expenditures, and past acquisitions. Track free cash flow, and judge each decision by how it changes per-share value, in both the numerator (total value) and the denominator (shares).
4. 4Ask 'what is the return?' on every proposal and go ahead only if the return is attractive under conservative assumptions. Reject projects that grow the numerator through overpaying for acquisitions or through capital projects that make no economic sense.
5. 5Split decision rights: delegate operations to the lowest, most local level, and keep capital allocation centralized with you. Munger called this an 'odd blend of decentralized operations and highly centralized capital allocation,' a mix that counters the institutional imperative.
6. 6Run the decision with minimal outside input. Do your own analysis (Stiritz arrived alone with a yellow legal pad; Buffett decided on an acquisition in a day without visiting the company). Treat bankers and advisers as optional, and keep your own counsel plus a select small circle.
7. 7Stay flexible and patient. There are no detailed strategic plans, because the right capital allocation varies with the situation. Hold cash or wait until an opportunity (an acquisition, or your own stock) clears your return bar, and act boldly only then.
8. 8Insulate yourself from the noise. Give no earnings guidance, skip Wall Street conferences, spend little time on investor relations, and keep a bare-bones office away from the financial center. Several outsiders were based in Denver, Omaha, St. Louis, or Los Angeles.

In Henry Singleton's Teledyne buyback, he bought back over 20 percent of shares outstanding. Because free cash flow was strong and interest rates had just dropped, he financed the whole repurchase with fixed-rate debt. Rates then rose and the bonds fell in price. Singleton did not believe rates would keep rising, so he retired the bonds, using cash from the company's pension fund, which was not taxed on investment gains. Teledyne financed a large repurchase cheaply, the pension fund booked sizable tax-free gains when rates later fell, and the stock compounded at over 40 percent a year for ten years. The move came from his own first-principles analysis of the financing and the tax treatment, not from imitation. His independence showed again in 1997, when Leon Cooperman asked about Fortune 500 companies announcing large buybacks and Singleton replied, 'If everyone's doing them, there must be something wrong with them.' He treated the popularity of the practice as a signal to be skeptical, and he repurchased shares only because the return was attractive in his own case.

• If peers are all doing the same thing (buybacks, acquisitions, guidance), then treat the consensus as a warning and re-derive the return yourself before joining.
• If a share repurchase is justified only by propping up the stock price or offsetting option grants, then reject it; buy back stock only when it is the best-returning investment available.
• If a project's return clears the hurdle only under optimistic assumptions, then do not proceed; require attractive returns using conservative assumptions.
• If a decision is operational and local, then push it down to the lowest level; if it is a capital allocation decision, then keep it with the CEO.
• If reported earnings and cash flow point in different directions, then follow cash flow, since net income is distorted by debt, taxes, capex, and acquisition history.

• Contrarianism for its own sake: you do the opposite of peers without running the numbers. The tell is that you can't state the expected return under conservative assumptions. The outsiders were rational first, and different as a result.
• Mistaking charisma for the job: you spend time on speeches, conferences, and press, and craft grand pronouncements. The tell is that your calendar is heavy with outward-facing events while capital allocation gets little attention. The outsiders skipped Davos and chamber of commerce speeches.
• Reliance on advisers and consensus: you hire bankers to validate deals and rely on their framing. The tell is that no one on your team can reproduce the deal logic on a yellow legal pad.
• Rigid strategic plans: you commit to a detailed multi-year plan and force capital into it. The tell is that you spend money because it is budgeted, not because the return is attractive now.
• Delegating capital allocation, or centralizing operations: you invert the loose-and-tight split. The tell is that operating managers decide on capital deployment, or the corporate staff second-guesses local operating calls.

• ▸Decision: describe the proposed investment, acquisition, buyback, or financing in one line.
• ▸Peer behavior: what are peers, analysts, and the press doing or saying about this? Is it driving your inclination?
• ▸Core economics: how does this change free cash flow per share, not reported net income?
• ▸Conservative-case return: the expected return using conservative assumptions, and the hurdle it must clear.
• ▸Alternatives: the best competing use of the capital, including buying back your own stock or doing nothing.
• ▸Rationale test: do you want to proceed for reasons of price support, option offset, imitation, or empire growth? Answer yes or no for each.
• ▸Outside input used: who advised you, and what did you decide independently of them?
• ▸Decision owner: confirm the CEO makes this capital allocation call; list what is delegated locally.
• ▸Verdict and date: proceed, wait, or reject, and the condition that would change it.

1. 1Map every dollar to the toolkit. List the five deployment choices (reinvest in existing operations, acquire a business, pay a dividend, pay down debt, repurchase stock) and the three funding sources (internal cash flow, debt, equity issuance). Each significant decision must be framed as a choice among these, not as a stand-alone project.
2. 2Size your own stake. Use Buffett's arithmetic: a company retaining earnings equal to 10 percent of net worth each year has, after ten years, deployed more than 60 percent of all capital at work in the business. Run this with your retention rate to see how much of the business your allocation decisions will have shaped. That number justifies keeping the decisions in your own hands.
3. 3Free your time by decentralizing operations. Concentrate operating authority in the general managers of the business units and keep headquarters thin. Singleton ran a wafer-thin corporate staff; Murphy's publishing head ran six dailies, magazine groups and weekly shoppers with three people at headquarters. Centralize capital decisions only because operating decisions are pushed down.
4. 4Hold a line-by-line annual budget session with each general manager. Follow Burke's model: each GM presents operating and capital budgets, and you and your CFO go through them in detail. Pay particular attention to capital expenditures and expenses. Expect material changes; a session that produces none is perfunctory. Use margins as a report card against peers.
5. 5Between sessions, leave managers alone. Burke's managers sometimes went months without hearing from corporate. The oversight sits in the allocation of capital and the budget review, not in daily operating involvement.
6. 6Judge each use of capital as an investor would. Compare value to price, and act boldly only when you see a compelling discrepancy. If you cannot find compelling projects, wait (the source says these CEOs were comfortable waiting, sometimes for long periods) rather than deploy capital to fill the calendar.
7. 7Treat your own stock as an investment option. When it is cheap, buy it, in large quantities if the discrepancy is large (Singleton eventually bought in over 90 percent of Teledyne's shares). When it is expensive, use it to buy other companies or to raise inexpensive capital for growth.
8. 8Do the thinking yourself and limit outside input. Do not hand allocation analysis to advisers or let Wall Street guidance steer it; outside interactions are distracting and time-consuming. Measure results by per share value and cash flow, not by growth, size or reported earnings.

Henry Singleton ran Teledyne for almost thirty years and spent most of his attention on deploying capital rather than managing operations. He focused capital on selective acquisitions and a series of large share repurchases, used debt frequently, issued shares sparingly, and paid no dividend until the late 1980s. His peers did the mirror image: issued shares to buy companies, paid dividends, avoided buybacks and used less debt. At the same time, Singleton ran an extremely decentralized organization with a wafer-thin headquarters staff, while his peers had elaborate staffs of vice presidents and MBAs. He also avoided analysts and journalists. The result was a 20.4 percent annual compound return; a dollar invested in 1963 was worth $180 by 1990, versus $27 for a broad group of conglomerates and $15 for the S&P 500. Capital Cities shows the same discipline in a different form: each general manager came to New York for budget meetings in which Burke and CFO Ron Doerfler went through operating and capital budgets line by line, and the sessions almost always produced material changes.

• If a capital request comes from a business unit, then decide it at the CEO level in the budget session; operating decisions stay with the general manager.
• If your stock trades well below your estimate of value, then compare repurchase against acquisitions and reinvestment first, and buy back aggressively if the discrepancy is compelling.
• If your stock is expensive relative to value, then use it as currency for acquisitions or to raise inexpensive capital, not for buybacks.
• If you cannot identify compelling projects, then hold the cash and wait; do not force deployment to match peers.
• If an outside adviser or Wall Street expectation is steering a decision, then re-derive the answer from cash flow and per share value yourself before acting.

• Delegating allocation to finance or business development by default. Tell: the CEO first sees a deal or buyback proposal when it arrives already packaged as a recommendation.
• Centralizing operations along with capital. Tell: headquarters staff grows with vice presidents and analysts, and general managers wait for approvals instead of running their units.
• Rubber-stamp budget reviews. Tell: sessions end with no material changes to capital expenditures or expenses.
• Following the institutional imperative: copying peers by paying dividends, avoiding buybacks or chasing size. Tell: your allocation mix looks like the industry average and you justify it by what competitors do.
• Measuring success by growth, size or reported earnings. Tell: allocation memos cite revenue or EPS impact but not cash flow or per share value.

• ▸Decision and amount: state the specific deployment being considered and its dollar size.
• ▸Toolkit category: reinvest, acquire, dividend, debt paydown or repurchase; and the funding source (internal cash flow, debt or equity).
• ▸Alternatives compared: list at least two other toolkit options and the expected return of each.
• ▸Value vs. price: your own estimate of value against the price paid or the current stock price, and the size of the discrepancy.
• ▸Cash flow basis: the cash flow (not reported earnings) the decision is expected to produce.
• ▸Per share impact: expected effect on long-term per share value, not on size or growth.
• ▸Own analysis: what you personally checked, and which outside advice or opinion you discounted and why.
• ▸Wait option: what happens if you do nothing, and what trigger would make the opportunity compelling.
• ▸Decision and date: your final call, the date, and when you will review the outcome.

1. 1Name your toolkit before you act. List the five uses of capital (reinvest in operations, acquire, pay dividends, pay down debt, repurchase stock) and the three sources (internal cash flow, debt, equity). Every decision is a choice among these, and the choice drives long-term shareholder returns as much as operating results do.
2. 2Define the standing default as waiting. Write down that when no option clears your return bar, you do nothing unusual. Dick Smith at General Cinema waited an entire decade without compelling projects. Inaction is a deliberate position, not an absence of strategy.
3. 3Build your own value estimate and compare it to price. These CEOs had confidence in their own analytical skills and acted only on rare, compelling gaps between value and price. Use cash-flow-based metrics (Malone popularized EBITDA as a pure measure of cash-generating ability) rather than the reported-earnings framing your peers use.
4. 4Pair each price signal with its matching tool. When your stock is cheap, buy it back, in large quantity if the gap is real. When it is expensive, use it as currency to buy a premium asset in a related field at a lower multiple, or to raise inexpensive capital. When a target is high-probability, use debt selectively and pay it down afterward.
5. 5Time actions to market dislocations. Malone and Magness repurchased stock during occasional market downturns to raise their stake. Keep dry capacity (cash flow, debt availability, a healthy balance sheet) so you can act when the dislocation arrives.
6. 6Concentrate on a few large, well-understood bets. Murphy did a small number of large acquisitions he knew to be high-probability, and moved slowly while building operational expertise. Do not substitute volume for conviction.
7. 7Decide on price, not attachment or ideology. Anders sold his favorite business, the F-16 unit, on the spot at $1.5 billion because it was the right per-share decision. Before any deal, ask whether the price beats your value estimate, and ignore how you feel about the asset.
8. 8After each bold action, return to the operating routine. Murphy's loop was to buy, improve operations, pay down debt, and repeat. Rebuild capacity before the next move rather than stacking deals.

Nick Chabraja at General Dynamics faced the Gulfstream acquisition. His predecessor Anders had avoided equity and leverage, relying on divestiture proceeds and healthy operating cash flow. Chabraja broke from that playbook and sold a lot of stock, which looks dilutive on its face. His reasoning was that the stock traded at 23 times next year's projected earnings against a historic average of 16. In his words: use a high-priced stock to buy a premium asset in a related field at a lower multiple and benefit from the arbitrage. Ray Lewis summarized it: Nick sold shares equal to one-third of the company to acquire a business that provided half of consolidated operating cash flow. The pattern is patience on equity issuance for years, then one large action when price and value diverged. Buffett's Gen Re purchase, paid for with stock at a record premium, follows the same logic.

• If your stock trades well below your estimate of value and you have no better use of cash, then repurchase aggressively and in size; if it trades at a significant premium to its historic multiple, then use it as acquisition currency or to raise capital.
• If no project clears your return bar, then wait, even for years, and do not lower the bar to stay busy.
• If a target is large, in a business you know, and a high-probability bet, then use debt selectively to buy it, and plan to pay the debt down through improved operations.
• If the price offered for something you own exceeds its value to you, then sell without hesitation, even if you love the asset.
• If an acquisition requires an unfamiliar industry, a fashionable diversification, or integration you cannot staff, then decline, however attractive the headline.

• Confusing activity with opportunism. You do many small deals to look decisive, as roll-ups did in the 1990s by acquiring too rapidly. The tell: integration lags acquisitions and debt climbs faster than operating improvement.
• Following fashion when waiting is right. You diversify into unrelated businesses and overpay for marquee properties, as Murphy's peers did. The tell: the deal rationale is about prestige or what rivals are doing, not a value-price gap.
• Treating one financing tool as doctrine. You never issue equity because your predecessor did not, and you miss the arbitrage when your stock sits at a record multiple. The tell: you cannot say why a tool is excluded in terms of price.
• Letting attachment override price. You keep a favored business when a buyer offers far more than it is worth to you. The tell: your reasons for holding refer to preference, not per-share value.
• Having no reserve when the moment arrives. You are leveraged or stretched when a downturn offers cheap stock or assets. The tell: a compelling opportunity appears and you cannot fund it.

• ▸Opportunity: describe the repurchase, acquisition, or issuance in one line.
• ▸Value estimate: your own intrinsic value or cash-flow-based figure, and how you derived it.
• ▸Price or multiple now vs. historic norm: for your stock or the target (e.g., 23x vs. 16x).
• ▸Alternatives compared: returns on reinvestment, debt paydown, dividend, repurchase, and acquisition.
• ▸Tool and funding source: which of the five uses and three sources fits the price signal.
• ▸Knowledge check: do you know this business well enough to call it high-probability? Yes/No and why.
• ▸Capacity: cash flow, debt headroom, and how you will pay down debt afterward.
• ▸Attachment test: would you make the same call if you had no emotional tie to the asset?
• ▸Decision: act now, wait, or decline, with the trigger that changes your answer.

1. 1Write the philosophy in one paragraph and print it where everyone sees it. Capital Cities put this paragraph on the inside cover of every annual report: 'Decentralization is the cornerstone of our philosophy. Our goal is to hire the best people we can and give them the responsibility and authority they need to perform their jobs. All decisions are made at the local level.' Include the cost-consciousness and sales-exploitation expectation, so autonomy arrives with a stated duty.
2. 2Name the autonomy holders. Concentrate operating responsibility and authority in the general managers of the business units (publishers, station managers). They hold the internal power and prestige. Headquarters does not.
3. 3Audit headquarters by function. List every corporate vice president and department, then test each against the Capital Cities headquarters: no functional VPs in marketing, strategic planning, or human resources; no corporate counsel; no public relations department. Murphy's secretary fielded all media calls. Cut or push down whatever fails the test. Define the surviving staff's purpose as supporting the general managers.
4. 4Remove layers, not just titles. At Chronicle Publishing, Sias and CFO Alan Nichols 'immediately eliminated an entire layer of executives at corporate headquarters.' Do this in the first move, not as a gradual program.
5. 5Pair autonomy with a rigorous budgeting process. Sias instituted one alongside the new authority. Set the numbers each manager owns. The rule at Capital Cities was that managers 'almost never heard from New York if they were hitting their numbers.' Silence from headquarters is the reward for hitting the budget.
6. 6Keep the few jobs that cannot be decentralized at the top. Murphy kept acquisitions as his own bailiwick, never delegating them or using investment bankers, while operations went to Burke and the field. Decide explicitly which allocation decisions stay central (capital allocation) and which go local (everything operating).
7. 7Expect and accept attrition. Chronicle general managers who were uncomfortable in the 'new, more demanding culture' left in the first year. Treat this as the structure selecting for independent, entrepreneurial managers, and do not backfill with people who need guidance.
8. 8Verify against results, not activity. Check unit margins against peers. Capital Cities had the highest margins in each of its business lines, and Chronicle's KRON margin went from 30% to 50%. If margins do not move, the autonomy is not working.

Chronicle Publishing in 1993 was a family-owned San Francisco media company torn by family squabbling, with operations suffering. John Sias, formerly head of Capital Cities' publishing division and the ABC Network, became CEO with a young CFO, Alan Nichols. They immediately eliminated an entire layer of executives at corporate headquarters, instituted a rigorous budgeting process, and gave significant authority and autonomy to the general managers. Many of those managers, uncomfortable with the more demanding culture, left in the first year. The results: margins at the NBC affiliate KRON rose 2,000 basis points, from 30 percent to 50 percent, and the Chronicle newspaper's operating margins more than doubled, from 4 percent to 10 percent. KRON was eventually sold for over $730 million in June 2000, and Hearst bought the paper for $660 million in 1999. Sias retired in 1999 after creating hundreds of millions of dollars of value for shareholders.

• If a headquarters role exists to originate strategy, marketing, or HR programs for the units, then eliminate it. If it exists to support general managers, keep it small.
• If a manager is hitting the budgeted numbers, then headquarters stays silent. Intervene only when the numbers slip.
• If you cannot name the budget a manager is accountable for, then do not grant autonomy yet. Build the budgeting process first or alongside it, as Sias did.
• If a decision allocates the company's capital (acquisitions, buybacks, divestitures), then keep it with the CEO. If it is operational, push it to the unit's general manager.
• If managers leave after the structure changes, then read it as selection and do not reverse course. Replace them with independent, entrepreneurial managers.

• Cutting staff but keeping control. Headquarters still approves local decisions, so autonomy is nominal. The tell: general managers hear from corporate regularly even when hitting their numbers.
• Autonomy without a budget discipline. Managers have authority but no hard numbers, so nothing triggers intervention or silence. The tell: no rigorous budgeting process and no cost-consciousness expectation in writing.
• Leaving a layer of executives in place while trimming staff. The tell: the organization chart still has intermediaries between the CEO and the general managers.
• Decentralizing capital allocation too. Acquisition and repurchase decisions drift into the units or to advisers. The tell: investment bankers drive deals rather than the CEO.
• Retreating after early attrition, or hiring managers who need direction. The tell: the first-year departures prompt a rebuilt central staff to compensate.

• ▸Philosophy paragraph: write the 3-4 sentence statement of autonomy and accountability to be published in the annual report.
• ▸Autonomy holders: list each unit general manager and the decisions they own fully.
• ▸Headquarters roles inventory: list every corporate function, VP, and department, with headcount.
• ▸Keep / cut / push-down verdict: for each role, mark whether it supports general managers (keep) or originates programs (cut or push down).
• ▸Layers to eliminate: name the executive layers between the CEO and the general managers to remove immediately.
• ▸Budget and numbers: state each unit's annual budget and the metrics that count as 'hitting your numbers.'
• ▸Centralized decisions: list the capital allocation decisions that stay with the CEO and who may not be involved (advisers, bankers).
• ▸Expected attrition: name the managers likely to leave under the new culture and the profile of their replacements.
• ▸Result check: record baseline and target unit margins versus peers, and the review date.

1. 1Name free cash flow per share as the scoreboard, and state in one sentence why reported net income is the wrong one. The passages say net income is 'a blunt instrument' distorted by debt levels, taxes, capital expenditures, and past acquisition history. Test it on your own company: list which of those four distortions move your reported earnings most.
2. 2Split the executive job into two tasks: run operations efficiently, and deploy the cash they generate. Assign an owner to each. Capital Cities did this with Dan Burke ('create the free cash flow') and Tom Murphy ('spend it'). If one person holds both roles, give the operations half to a strong COO.
3. 3Put every use of cash on one list drawn from the five deployment choices: invest in existing operations, acquire businesses, pay dividends, pay down debt, repurchase stock. Add the three funding sources: internal cash flow, debt, equity. Rank the options by expected return to shareholders, not by their effect on this quarter's EPS.
4. 4Evaluate each investment decision on the cash it returns, then compare it against the alternatives on the list. A share repurchase competes directly with an acquisition. Singleton at Teledyne used selective acquisitions and large repurchases, used debt frequently, issued shares sparingly, and paid no dividend until the late 1980s. His peers did the reverse.
5. 5Make the cost base the controllable lever. Murphy and Burke noted that you cannot control revenues at a TV station but you can control costs. Set cost consciousness as an explicit expectation of operating managers, and cut headquarters overhead to a thin staff whose job is to support the general managers.
6. 6Push operating authority to the unit managers who generate the cash. Capital Cities printed the policy on the inside cover of every annual report: 'All decisions are made at the local level.' Managers who hit their numbers are left alone. Stop requiring reports that do not change decisions. Burke dropped his weekly memos to Murphy after months without a response.
7. 7Align pay and communication with the cash focus. Keep Wall Street communication occasional and factual rather than frequent and promotional. The cash-focused CEOs avoided Davos, books of management advice, and chamber of commerce speeches.
8. 8Track the result as per-share value growth relative to peers and the market. Do not track the absolute rate of return, revenue growth, or earnings growth.

Capital Cities under Tom Murphy shows the construct operating end to end. Murphy became CEO at forty, when the company had $28 million in revenue, and promoted Dan Burke to president and COO. Burke ran daily operations and 'create[d] the free cash flow,' and Murphy handled acquisitions, capital allocation, and occasional Wall Street contact. Murphy deployed that cash into a run of acquisitions: KTRK Houston for $22 million in 1967, Fairchild Communications for $42 million in 1968, Triangle Communications for $120 million in 1970, the Fort Worth Telegram for $75 million in 1974, the Kansas City Star for $95 million in 1977, and Cablecom for $139 million in 1980. In the bear market of the mid-1970s to early 1980s he switched tools and bought back close to 50 percent of the shares, most of them at single-digit price-to-earnings multiples. Headquarters had no vice presidents for marketing, planning, or HR, no corporate counsel, and no PR department. Managers who hit their numbers rarely heard from New York. The result is the pattern the passages describe: cash generated by cost-conscious local managers, then allocated by one person to the highest-return option available.

• If a choice would raise reported EPS but lower free cash flow per share, reject it. The cash-focused CEOs optimized free cash flow, not quarterly net income.
• If your shares trade at low multiples and no acquisition clears your return bar, buy back stock. Murphy bought in close to 50 percent, mostly at single-digit P/Es, during a bear market.
• If a headquarters function does not help a general manager hit their numbers, eliminate it. The model is a thin corporate staff and local decision rights.
• If revenues are outside your control, target costs. Costs are the controllable side of the cash equation.
• If you cannot both run operations and allocate capital well, split the roles. Pair the CEO with a COO who owns the daily operations.

• Judging performance on revenue or earnings growth because the press and Wall Street do. The tell: your internal scorecard has no per-share value or relative-return measure.
• Treating capital allocation as an afterthought to operations. The tell: the board reviews operating plans in detail, but no one can state the ranked alternatives for deploying this year's cash.
• Making each investment decision in isolation. The tell: acquisitions get approved without being compared to repurchases or debt paydown, and the five-option list never appears.
• Defaulting to peer-style allocation: issuing shares for deals, paying dividends, avoiding buybacks, and using little debt. The tell: your allocation pattern matches your peers' and your results do too.
• Building a large headquarters and reporting apparatus that slows the managers producing the cash. The tell: field managers spend their time on reports to headquarters, and costs and friction rise.

• ▸Business unit and period: name the unit and the year being evaluated.
• ▸Reported EPS: record the net income per share figure Wall Street sees.
• ▸Distortions to EPS: note the effect of debt level, taxes, capital expenditures, and past acquisitions on that figure.
• ▸Free cash flow per share: record the cash figure you are optimizing and its change from last period.
• ▸Cash available to deploy: state the total from internal cash flow, debt, and equity, with the funding source for each.
• ▸Ranked deployment options: list the five choices (existing operations, acquisitions, dividends, debt paydown, repurchases) with the expected return of each.
• ▸Decision and rationale: state which tool you chose and why it beats the alternatives on return to shareholders.
• ▸Cost control actions: list the specific cost and overhead items the unit manager controls and the target for each.
• ▸Relative result: record per-share value change versus peers and the market.