The CEO Portfolio Manager | The Consilient Investor 05
How Singleton created a multi-decade 20% CAGR success story, and how Eddie Lampert ran the same playbook into bankruptcy.
As thanks for your time, this post includes a FREE downloadable PDF scorecard you can apply to your next analysis.
In the mid-1970s, while his Fortune 500 peers were chasing the cover of Forbes with bigger acquisitions and bolder forecasts, Henry Singleton was doing something almost no other CEO would do: he was buying back his own stock.
Not a small program. Not a quiet, formulaic repurchase plan. Eight tender offers between 1972 and 1984 that systematically retired roughly 90% of Teledyne’s outstanding shares.
“In October, 1972, we tendered for one million shares and 8.9 million came in. We took them all at $20 and figured it was a fluke, and that we couldn’t do it again. But instead of going up, our stock went down.
So we kept tendering, first at $14 and then doing two bonds-for-stock swaps. Every time one tender was over the stock would go down and we’d tender again, and we’d get a new deluge. Then two more tenders at $18 and $40” - Henry Singleton
His logic was unfashionable and largely ignored at the time. When the stock was cheap relative to its underlying earning power, the best use of every available dollar was to retire it.
When the stock was expensive - when it traded above what Singleton thought the business was actually worth - he reversed direction and used Teledyne shares as currency to buy other businesses.
Singleton was not allocating capital like he was running a conglomerate, he was allocating capital like he was running a portfolio. It just so happened that his publicly traded operating company was the investent vehicle.
The result of his capital allocation was extraordinary - Teledyne compounded at a compound annual growth rate (CAGR) of 20.4% over two decades.
Capital allocation is the most distinct responsibility of a CEO, yet most rise through sales or operations and are ill-equipped for this finance-heavy task.
In The Consilient Investor 02 we looked at the mathematics behind value creation. In this fifth installment of the series we’ll connect those threads to the human being whose decisions actually determine where the cash flow goes.
We’ll walk through the nine allocation levers a CEO can pull, examine the archetype of the great capital allocator, the archetype of the destructive one, and end with five principles you can use to separate the two.
Missed a previous issue of The Consilient Investor? Catch up on them here:
The Job Nobody Trains For
If operations are how the company makes money, then capital allocation decides what to do with that money.
Almost nothing else they do - running operations, motivating employees, navigating regulators - is unique to the role. Capital allocation is the most distinct CEO responsibility, and yet many do it poorly.
Most public-company CEOs rose through sales, marketing, operations, or engineering. Very few rose through corporate finance. And yet, the moment they take the corner office, they are handed a check the size of their company’s annual free cash flow and asked to allocate it across nine competing claims - without, in most cases, ever having been asked to think rigorously about per-share value.
“Once they become CEOs, they face new responsibilities. They now must make capital allocation decisions, a critical job that they may have never tackled and that is not easily mastered.” - Warren Buffett
The result is that long-term shareholder returns are determined as much by what management does with operating cash flow as by the operating cash flow itself. Two companies with identical underlying economics can produce wildly divergent outcomes for shareholders if one is run by a Singleton and the other is run by someone who believes the job is to grow the empire.
The Nine Levers of Capital Deployment
Every dollar that arrives from a source of capital can be put to work in nine possible ways.
Each application has its own internal rate of return, its own optionality, its own embedded message to the market, and its own tax treatment. The CEO’s job then is to decide, each quarter and each year, which lever to pull.
“The answer to nearly every capital allocation question is, ‘it depends.’” - Mauboussin and Callahan, Capital Allocation: Returns, Analysis and Assessment (2025)
1. Reinvesting in the business
Capital Expenditures (CapEx)
CapEx is investing in physical assets - plants, equipment, locations. This is the most tangible form of reinvestment. You can walk through it. This can include both maintenance CapEx and growth CapEx.
As the developed world economy moves towards intangibles (services, software, IP), the overall amount of CapEx spend - as a percentage of revenue - is declining, although this trend has reversed recently with the Hyperscaler AI infrastructure buildout.
Investing too little in CapEx can starve the business of necessary tangible assets and hurt future returns, while investing too much can destroy value. The question then isn’t “should we invest in CapEx?” but “what is the marginal return on this incremental dollar of CapEx, and is it above our cost of capital?”
Once a CEO commits to a multi-year capacity expansion, there is enormous institutional pressure to maintain it regardless of the actual incremental returns. The cement industry, the airline industry, and large parts of telecom are graveyards of CapEx programs that were green-lit because the project was “strategic,” even though the marginal ROIC sat below the cost of capital.
Working Capital
Working capital is funding inventory and receivables. For capital-intensive growth businesses this can consume more cash than CapEx. For example, an aircraft-manufacturer like Boeing or Airbus will have massive amounts of materials and components in inventory, coupled with long production cycles and delayed receivables, as final payouts happen upon delivery.
Working capital is significant because it ties up cash and can harm liquidity - this is exceptionally important during periods of rapid market shifts (e.g. rapid growth, or rapid obsolescence of inventory).
When working capital balloons faster than revenue it is almost always a sign that something is wrong. Either demand is softening (inventory is piling up), customers are paying more slowly (receivables are stretching), or both.
Working capital can also be a deliberate offensive investment - building inventory ahead of an anticipated demand cycle, extending favorable credit terms to win share from undercapitalized competitors, or pre-stocking long-lead components to insulate the supply chain.
The best read on a CEO's working-capital discipline is the cash conversion cycle.
A negative CCC, like Costco's, Hermès', or Apple’s, means the business is being funded by its suppliers and customers rather than its own shareholders. This is the structural equivalent of a permanent, interest-free loan, and one of the most under-appreciated forms of capital allocation.
Mergers & Acquisitions (M&A)
M&A is buying growth. This is the lever with the most variance, both up and down. The base rates on most M&A is - depending on the study, somewhere between half and three-quarters of large deals destroy value for the acquirer’s shareholders.
“One study of 1,267 deals from 1995 to 2018 showed that the stock price of the buyer went down 60 percent of the time upon announcement and that the average change for the full sample was -1.6 percent. These aggregates naturally hide a lot of variance. Plenty of transactions create value for the buyers.” - Mauboussin and Callahan, Capital Allocation: Returns, Analysis and Assessment (2025)
A small number of exceptional serial acquirers - Constellation Software, HEICO, Berkshire - have built decades-long track records of disciplined, small, programmatic acquisitions at attractive multiples.
The worst M&A deals tend to share a specific pattern: a transformative, large-scale acquisition done at the peak of a cycle, marketed to shareholders as “strategic” rather than financially justified, and accompanied by a synergy estimate that nobody on the buy side believes. AOL Time Warner is the prototypical example, and SpaceX acquiring xAI is a real-time example.
The investor then must track ROIC on the cumulative dollars deployed in acquisitions (purchase price plus goodwill plus integration costs) over a five-to-ten-year window, and compare it to the cost of capital. Also look for any adjustments of acquisition-related expenses or intangibles that management may make to flatter EPS - this is the accounting recognition of overpayment.
Research & Development (R&D)
R&D is the investment into intangible innovation. R&D is expensed on the income statement the period it is incurred, even though the economic life of the resulting intellectual property may be measured in decades. Thus, GAAP systematically understates the earning power of R&D-heavy companies because of this asymmetry.
R&D-heavy companies may appear overpriced on a P/E basis, as the value captured by strong intangible reinvestment programs is not captured (e.g. Costco’s labor investment, Ferrari’s engine development, Hermès’ atelier and apprenticeship system).
As a percentage of revenue, R&D spend has been growing with time. This correlates with the broader shift from investment in tangible assets (CapEx) to intangible assets.
R&D is also the lever most prone to measurement failure. There is no straightforward way to compute the ROIC of an R&D dollar. The best you can usually do is look at long-run revenue or operating-profit growth per dollar of cumulative R&D and compare it across competitors.
For example, Meta’s Reality Labs spend makes this calculation almost impossible in real time; the entire value proposition of this spend hinges on whether or not the long-run ROIC will eventually clear the cost of capital.
Selling, General and Administration (SG&A)
SG&A is investing in the intangible infrastructure of the business - brand-building, customer acquisition, sales-force training, organizational capability, distribution and relationships. Like R&D, GAAP runs this through the income statement as a period expense.
Like R&D, a business with high investment SG&A looks more expensive on a P/E basis than its underlying economics warrant, because GAAP earnings systematically understate cash earning power.
But economically, a dollar spent training a Hermès sales associate, building HEICO’s aftermarket-parts distribution relationships, or selling to a SaaS customer with a five-year payback is no different from a dollar spent on a factory or a server farm. The consilient investor views this as an investment, accounted for as an expense, as it creates intangible assets.
However SG&A is also open to interpretation. When management cuts SG&A to defend margins in a soft quarter, they may be cutting waste, or they may be cutting the very investments that drive future returns.
It is very hard to distinguish from the outside. The cleanest tell is whether SG&A as a percentage of revenue tracks in a stable band (consistent with disciplined reinvestment) or whether it expands and contracts with the calendar (consistent with budget management dressed up as strategy).
2. Returning Capital to Shareholders
Share Buybacks
“Buybacks are divisive. They divide people who do understand finance from people who don’t.” - Kenneth R. French, Professor of Finance, Tuck School of Business at Dartmouth College.
Repurchasing equity is often one of the most misunderstood - and misused - levers in the entire framework. As Singleton showed with his pioneering approach, when done correctly, buybacks compound per-share value at extraordinary rates.
However when done incorrectly, they amount to a wealth transfer from one group of shareholders to another.
Mauboussin and Callahan write:
“To be clear, buying back undervalued or overvalued stock does not create or destroy value for the company. It transfers wealth from one group of shareholders to another.
A company that buys back undervalued stock transfers wealth from the sellers to the ongoing holders.
A company that buys back overvalued stock transfers value from the ongoing holders to the sellers.”
Most CFOs will state that they only buy back shares when the shares are undervalued, although in reality they are often executed poorly.
“Fifty to eighty percent of management will say the stock of their company is undervalued when surveyed in a typical quarter. The most common method that CFOs cite for valuing the stock of their company is “current price relative to historic highs and lows.” This does not instill confidence.” - Mauboussin and Callahan, Capital Allocation: Returns, Analysis and Assessment (2025)
There’s a measurement nuance worth commenting on: time-weighted vs. dollar-weighted returns. Time-weighted return is the movement of the share price over time, while dollar-weighted return is what shareholders actually earned, adjusted for when capital came in (issuance) and when it left (buybacks). The two only match if all equity transactions happened at fair value.
When CEOs issue stock cheap or buy it back rich, dollar-weighted returns fall below time-weighted returns - even if the stock chart looks great.
This is the hidden cost of bad buyback timing and dilutive stock-based comp (SBC). Mauboussin’s data shows that companies in the top quintile of SBC-as-a-percent-of-sales produce lower future excess returns than companies in the bottom quintile.
Total SBC for U.S. public companies grew from $26 billion in 2006 (0.2% of sales) to $290 billion in 2023 (1.3% of sales). This is a structural wealth transfer from outside shareholders to insiders, often masked by gross buybacks that just offset the dilution.
When you see a company aggressively buying back stock, always check the net share count, not just the gross repurchase number. A company that retires $5 billion of stock while issuing $4.5 billion in SBC isn’t really returning capital to you — it’s running on a treadmill at your expense.
Friend and fellow Substack author (Undiscovered Compounders) proposes the following framework for assessing the rationality of buybacks:
Dividends
Dividends are cold, hard cash payments to equity owners - the most direct form of capital return. They may be taken as a sign that the company has more cash than it can productively reinvest at its cost of capital. They may be seen as an admission that the reinvestment runway has shortened.
However dividends also impose discipline. They are a public commitment, harder to cut than to declare, and they force management to confront the question of whether retained capital is actually being deployed at returns above what shareholders could achieve on their own. Because of this, dividends tend to have less volatility relative to buybacks.
However, due to tax implications and the rise of buybacks, dividends are becoming less prominent with time.
3. Improve Balance Sheet
Debt Repayment
Repaying debt is the most conservative lever, with a guaranteed return that is the after-tax cost of the debt itself. For most investment-grade companies, that’s a low single-digit number - well below the cost of equity, often below the long-run return of the underlying business.
Thus on the surface it appears to be one of the least desirable forms of capital allocation. However debt often introduces fragility, and having balance sheet flexibility increases optionality, and so this lever can be a rational choice.
Those companies with the best balance sheets - the lowest amount of debt - tend to also have the highest return on invested capital (ROIC).
Holding Cash
When management is indecisive, or none of the other eight levers look attractive enough, then cash piles up on the balance sheet.
Many companies chose to build cash, and at the end of 2024 U.S. public companies held roughly $2.5 trillion in cash. The distribution is wildly skewed: ten companies hold about a quarter of all the excess cash in the system, and Berkshire alone sits on nearly $400 billion at time of writing - more than the GDP of most countries.
As the economy has shifted from tangible to intangible investment, intangibles change the value of cash. R&D, brand, software, and human capital make terrible collateral - you can’t pledge an intangible asset the way you can pledge a steel mill.
That means intangible-heavy companies have less debt capacity and need more cash on hand to self-fund their reinvestment. The correlation between SG&A intensity and cash holdings is 0.76. Healthcare companies hold 39% of their assets in cash; tech holds 23%; utilities hold less than 1%.
The trade off is that cash earns below the cost of capital. But cash also buys something that isn’t easily quantifiable: optionality. The best deployment opportunities - counter-cyclical buybacks, distressed acquisitions, share gains during a competitor’s blow-up - almost always show up when capital is scarce for everyone else. This is likely why companies with low debt have higher ROIC - they have more optionality.
“[Warren Buffett] thinks of cash differently than conventional investors. He thinks of cash as a call option with no expiration date, an option on every asset class, with no strike price.” - Alice Schroeder, author of The Snowball: Warren Buffet and the Business of Life.
Thus, idle cash is only a problem if the CEO isn’t intentional with the tradeoff they are making. If the intention is “we’re a maturity-stage compounder with low reinvestment needs and we’d rather not destroy value through bad M&A or buybacks at peak multiples” - then that’s discipline.
If the intention isn’t clear to the CEO, then a growing cash pile is going to be a drag on returns.
Doing nothing is a choice, and it should be made for the same reason as every other choice on the list: because it’s the highest expected value use of the next marginal dollar.
In short, the CEO’s job is to keep a constantly updated mental ranking of the marginal IRR available through each of these nine levers and to allocate the next dollar to the opportunity with the highest risk-adjusted return. This is why the answer is always “it depends”.
With that covered, let’s now examine the track record of two prominent capital allocators.
Singleton: The Archetype
The introduction told you what Singleton did. The more important question is how he did it.
Singleton ran Teledyne from 1960 to 1991. He held a doctorate in electrical engineering from MIT, and not a single accounting class to his name. Almost no one with his background - and almost no one in that era, save for Warren Buffett - would have thought of himself as a portfolio manager for his own shareholders.
But that is precisely what he was.
Through Teledyne, Singleton was running a portfolio of operating cash flows and his overarching goal was to compound per-share intrinsic value.
Each lever was an action he could take in managing that portfolio, and his job was to act based on each available IRR. When the public market mispriced Teledyne, the highest-IRR position was retiring his own stock. When other operating businesses came up cheap, the highest-IRR position was buying them. When neither was true, the best move was patience.
“I know a lot of people have very strong and definite plans that they’ve worked out on all kinds of things, but we’re subject to a tremendous number of outside influences and the vast majority of them cannot be predicted. So my idea is to stay flexible.” - Henry Singleton
The most heretical position he took was on dividends. At a time when roughly 70% of public companies paid them, Teledyne paid almost none. Singleton’s logic was unsentimental: a dollar retained inside Teledyne and redeployed at a high ROIC - or used to repurchase undervalued shares - was worth more than the same dollar handed to a shareholder, taxed, and reinvested elsewhere.
The long-term results validated his thesis.
But Singleton’s playbook is not universally applicable, and this is the detail his imitators miss. It worked because Teledyne owned a portfolio of high-ROIC operating businesses, and because he only bought back periodically when they were obviously accretive.
Strip either condition away - replace the high-ROIC engine with a declining one, or add blanket buybacks indiscriminate of value - and the same set of moves stops compounding value and starts destroying it.
Which brings us to a man (and his public market investors) who learned this the expensive way.
Lampert: The Antithesis
Eddie Lampert came up through Goldman Sachs and started his own hedge fund, ESL Investments, in his late twenties. He took control of Kmart out of bankruptcy in 2003, merged it with Sears in 2005 to form Sears Holdings (valued at the time at roughly $11B), and installed himself as chairman and eventually CEO.
The framing was that he was a financial-discipline maestro conducting turnarounds in an underinvested retail empire, and that he would apply the Buffett-Singleton playbook to extract enormous shareholder value.
Sears Holdings repurchased roughly $5.8B of stock between 2005 and 2010 at an average price around $89 per share. That same stock subsequently went to zero, with the company filing for Chapter 11 in October 2018.
Valuable operating assets - Lands’ End, Sears Hometown, Orchard Supply Hardware, and the Craftsman brand - were spun off or sold piecewise between 2011 and 2017. As Peter Lynch would say, this is cutting the flowers to water the weeds.
In 2015, hundreds of Sears’ most attractive real estate properties were transferred for $2.7B to Seritage Growth Properties, a REIT that Lampert separately chaired and substantially owned - a related-party transaction that became the subject of significant subsequent litigation.
The namesake store count fell from thousands at the Kmart-Sears merger to fewer than 700 at the bankruptcy filing, and to a single-digit handful today.
On the surface, Lampert’s actions are the Singleton playbook applied to retail: aggressive buybacks, refusal to over-invest in low-return operations, and monetization of trapped balance-sheet value.
However, Singleton ran a portfolio of high-ROIC operating businesses and used buybacks as an opportunistic supplement when his own stock was demonstrably cheap relative to those businesses’ earning power.
Lampert ran a declining low-ROIC operating business and used buybacks to compensate for the underlying decline. Further, this was funded by the systematic liquidation of the only assets - the real estate, the brands, the better-performing subsidiaries - that gave the operating company any residual value.
CapEx was systematically slashed for over a decade; store-level reinvestment in fixtures, inventory, and digital infrastructure was a fraction of what competitors (Walmart, Home Depot, Target) were spending per square foot through the same period. The business was being starved of reinvestment to fund buybacks.
Lampert’s defenders argued, and still argue, that the underlying retail business was already terminal and that the buybacks and asset sales were a rational extraction of value from a declining enterprise.
While this is likely true, the rational capital allocation move was not to buy back the holding company’s stock at $89 - it is to return capital to shareholders in the cleanest available form (dividend or wind-down), let them redeploy it to higher-return uses elsewhere, and avoid the wealth transfer that pro-cyclical buybacks create. Lampert did the opposite.
The bitter lesson here is that the Singleton archetype is dangerous when imitated in form, without substance. Buybacks alone are not a strategy; they are the byproduct of a much larger set of decisions about ROIC, reinvestment runway, and per-share intrinsic value.
Do you have a Singleton or a Lampert?
“It’s who you are, not who you say you are. It’s what you do, not what you say you’re going to do.” - Mark Miller, President of CSI
James McTaggart (et al) described four principles of resource allocation in their book, The Value Imperative (1994), and Mauboussin adds a fifth regarding a bias to action.
These principles can provide a framework to measure management’s mindset regarding capital allocation.
It is insufficient to rely on surface-level metrics such as Earnings Per Share (EPS), Revenue Growth, or Return on Equity (ROE). These metrics are susceptible to the distortions of leverage, share buybacks, and accounting conventions that divorce reported figures from cash-flow realities.
An increase in EPS, for instance, can be manufactured through debt-funded buybacks even as the underlying economic engine of the business deteriorates.
Similarly, revenue growth is not inherently virtuous; growth is merely an amplifier of the underlying economic physics of the business.
The gold standard quantitative metric for value creation is ROIC and ROIIC.
“Management’s goal should be to create value by making investments that earn a return in excess of the opportunity cost of capital.” - Mauboussin and Callahan, Capital Allocation: Returns, Analysis and Assessment (2025)
You now have the framework - try it with a company in your portfolio!
This FREE PDF worksheet walks you through scoring all nine levers and the five principles of capital allocation.
It’s free - grab it here. 👇
What This Means for The Consilient Investor
To most of Wall Street at the time, Henry Singleton’s tender offers looked like the desperate moves of a CEO who didn’t know what else to do. Analysts downgraded the stock. The financial press wouldn’t put him on the cover.
He kept buying anyway.
Two decades later, CEOs continually try to implement his playbook and most of them, including the ones still trying today, are missing what made it work.
The lesson isn’t that buybacks compound shareholder value. The lesson is that the CEO’s job is to keep a constantly updated mental ranking of the marginal IRR available across the nine levers and to allocate the next dollar accordingly. Singleton did this.
Three things follow.
Firstly, the quality of a business and the quality of its capital allocator are two separate variables. A great business in the hands of a poor allocator produces mediocre per-share returns; a merely good business in the hands of a great allocator can produce extraordinary ones. This is why the five principles of capital allocation matter.
Secondly, the nine levers are not interchangeable. A CEO who repurchases stock every quarter regardless of valuation is not exercising judgment, they are running a formula. A CEO who suspends the program when the stock gets expensive and resumes when it gets cheap is doing the actual job.
Finally, Capital allocation is the most legible window into management quality you have as an outside investor. You cannot easily evaluate operational excellence from a 10-K. You can evaluate capital allocation. Pull the share-count history. Trace cash from the cash flow statement to its nine destinations. Calculate ROIC and ROIIC over rolling five-year windows. Compare the M&A track record to the cost of capital. The ledger is the report card.
So when you read the next earnings release or listen to the next conference call, ignore the headline numbers.
Use the PDF worksheet above and assess the five characteristics. Ask which of the nine levers management pulled - and whether it was the highest-IRR lever available to them. Track the ROIC. Assess whether the pattern of decisions actually created per-share value.
If the answer is yes - consistently, over many years, through many cycles - you have a Singleton. Hold tight, and let the compounding do its work.
If the answer is no, then you are looking at managers who have not yet realized that the most important job they have is the one nobody trained them for.
Missed a previous issue of The Consilient Investor? Catch up on them here:
Looking for a new read? Continue the pursuit! Check out The Library Below 📚
Further Reading:
The Outsiders: Eight Unconventional CEOs and Their Radically Rational Blueprint for Success. William N. Thorndike JR. 2012.
The Snowball: Warren Buffett and the Business of Life. Alice Schroeder. 2009.
Capital Allocation: Results, Analysis and Assessment. Michael J. Mauboussin and Dan Callahan. Counterpoint Global Analysis. (2025)
Cash Holdings: Data, Theory and Alternatives. Michael J. Mauboussin and Dan Callahan. Counterpoint Global Analysis (2025)
Which One Is it? Equity Issuance and Retirement. Michael J. Mauboussin and Dan Callahan. Counterpoint Global Analysis (2024)
Capital Allocation is The Art Of Optimizing The EPS & ROA.
.
For every new project, Singleton demands 20% of ROA.
.
Singleton looks at deeper level of efficiency at ROA, not at the shallower ROIC.
Lampert in contrast to Singleton is excellent; thanks for ferreting that one.