“Most businesses consume cash. Very few ever create it for the security holder on a regular basis.”
— Martin J. Whitman, founder of Third Avenue Management
Download the Slide Deck (PDF) : Direction_of_Absorption
Listen to the deep-dive discussion – The Direction of Capital Absorption
I. The Flow Between Two Reservoirs
Every business is a flow of capital between two reservoirs: the toll it collects today, and the growth it funds for tomorrow. The A and the B. In an earlier essay, we examined a peculiar case — a business so profitable that its toll overflowed. Visa generates cash at thirty percent on capital, but cannot reinvest it at that rate; the network is built, and the next transaction costs nothing to process. The surplus had to leave, through dividends and buybacks. We called the gap between what it generates and what it can absorb the recycling spread, and we called it a beautiful problem — a problem of wealth.
This essay is about the direction of that flow. Not how much capital moves, but which way it runs. Whether the toll absorbs the growth, or the growth absorbs the toll.
It is the most fundamental property of a business — more fundamental than how fast it grows, or even whether that growth creates value — because the direction of absorption is what those things are made of.
II. A + B versus A × B
The framework has always drawn a line between two structures. In the additive business — A + B — the toll and the growth are separate, and the growth is funded out of the toll. The B is a line item the A pays for. Most businesses live here: the cash from today’s operations is poured back into tomorrow’s expansion, and what matters is whether that expansion earns more than it costs.
In the multiplicative business — A × B — the toll and the growth are fused. The wave is not a line item the toll funds; it arrives from outside, at no cost, and the toll simply stands in its path and widens. Visa does not pay for the world to digitize. The A overflows the B structurally, because the B costs almost nothing.
The distinction is not cosmetic. It determines the direction in which capital is absorbed — and that direction, as we will see, governs everything from how much cash returns to shareholders to whether the business can be destroyed at all. But absorption comes first. The destruction is only its furthest consequence.
III. The Direction of Absorption
Here is the central idea. In a business of quality — whether its tolls are heavy or light — the A absorbs the B. The toll generates more than the growth consumes, and the excess overflows. In a poor business, the flow runs the other way: the B absorbs the A. The growth is so costly that it devours the cash the toll produces, and reaches for more.
Notice what this is not. It is not a distinction between Freesurfers and everything else. A heavy toll — a railway, a pipeline, an aircraft-engine franchise — can have the A firmly absorbing the B, the toll funding its maintenance and expansion and still throwing off cash. It is on the right side of the flow without being a Freesurfer at all. The dividing line is not the category of business. It is the direction of absorption, and it cuts across every category.
Quality is not a matter of how fast a business grows, or how large its moat looks. It is a matter of which way the capital flows between the toll and the growth. That single direction is the signature.
IV. Dual Duration as the Dial
Almost every business carries a dual duration — a short-duration component (the cash the toll returns now) and a long-duration component (the growth compounding for later). The exception is the pure B that never had an A: a business that is all growth and no toll, funded entirely from outside capital. Everywhere else, the two durations coexist, at varying strength.
The dual duration is not created by the surplus. It is the dial that reveals the direction of the flow. When the A overflows the B, the dial points short: cash is returned to shareholders. When the B begins to absorb the A, the dial weakens — the short component evaporates, and the business shifts from returning surplus to funding the growth entirely, until there is nothing left to return.
At one extreme stands the pure A — the magnificent toll with almost no B. See’s Candies. Coca-Cola. The classic Stalwart: limited growth, but enormous, unabsorbable surplus. The A so dwarfs the tiny B that it overflows massively. The dial points almost entirely short — nearly all the cash is returned, because there is almost no growth to fund. This is return of capital in its purest form. Munger had to ship See’s cash to Omaha precisely because the toll overflowed so completely that it had to be redeployed elsewhere.
The Freesurfer sits one step further along. A magnificent toll, but with a free B that grows. The A overflows and the B compounds, at no cost. The dial points both ways at once: surplus returned now, and free growth compounding for later. The pure A overflows but does not compound; the Freesurfer overflows and compounds. Both are on the right side of the flow. The Freesurfer simply adds an engine to the overflow.
Read the dial, and you read the direction. A strong short component is overflow. A vanishing one is absorption. The dual duration does not cause the flow — it displays it.
V. The Three Zones
The flow is not binary. Between the toll that overflows and the growth that devours lies a spectrum with three zones.
Zone One — the A absorbs the B. The toll overflows. The pure A (See’s), the Freesurfer (Visa), and the healthy mechanical compounder (a mature cloud business reinvesting above its cost of capital) all live here. Value is created, and capital returns to the owner. The flow runs in the right direction.
Zone Two — a good toll funds a value-destroying B. This is the subtle case, and the most common. A business has one or two genuine tolls — sound A’s that overflow — but redirects that surplus into a B that never clears its cost of capital. The company does not die; the healthy tolls keep it alive. But the global value erodes, because the cash the good tolls produce is siphoned into growth that destroys more than it creates. The business is worth less than the sum of its sound tolls — a conglomerate whose strongest division quietly subsidizes its weakest, a franchise pouring its winnings into an unmoated venture that consumes more than it returns.
Zone Three — the B absorbs the A entirely. No sound toll remains to carry the company. The costly growth devours the toll and reaches for outside capital to survive. The pure B — growth with no toll at all — sits here. So does any business whose every toll has been consumed by an insatiable B.
Most real destruction of value is not Zone Three. It is Zone Two — not the company that fails, but the good business that erodes a part of itself by funding the wrong absorption.
VI. Temporary versus Permanent
A business whose B absorbs its A is not automatically doomed. The absorption can be a phase. Every investment phase looks, for a time, like the growth devouring the toll: the mature A funds an emerging B that does not yet pay for itself. Amazon’s retail toll funded the construction of AWS for years before AWS overflowed on its own.
What separates a phase from a trap is a single condition: the reversal must be credible. At some point the B must mature enough that it begins to overflow — that the A reabsorbs it, and the flow turns back to the right direction. Value is destroyed not when the B absorbs the A, but when that reversal never arrives, or was never plausible. The investment phase and the value trap are the same picture in the present tense. The difference lies entirely in whether the absorption is on its way to reversing, or drifting with no overflow ever in sight.
VII. What Finance Already Knew — and Never Connected
Each piece of this has been modeled before. The pieces have simply never been joined.
The theory of the absorbing barrier is nearly seventy years old. De Finetti, in 1957, modeled a company’s cash reserves as a random walk with zero as an absorbing state — a point of no return — and asked for the dividend strategy that maximizes payouts before ruin. An entire actuarial literature refined it. But it lives in insurance and stochastic control, applied to the probability of an insurer’s failure, never to the quality of a business or the direction of its capital flow.
The threshold of value-creating growth is just as well established. The value-driver models of corporate finance state it cleanly: the cash available to owners is what remains of operating profit after the reinvestment required to grow, and growth adds value only when the return on that reinvestment exceeds the cost of capital. But this is static valuation — it measures whether growth creates or destroys value at an instant, and never speaks of absorbing barriers, of ruin, or of the direction of the flow through time.
And the surplus problem has its own literature — free cash flow theory, the discretion that comes from generating more cash than the business can absorb. But it is treated as a question of how that surplus might be wasted, never as the signature of which way the capital flows.
Finance has the absorbing barrier, the reinvestment threshold, and the surplus, each in its own sealed room. The direction of absorption is the corridor that connects them. The reinvestment threshold determines which way the flow runs; the direction of the flow determines the distance to the absorbing barrier; and the surplus is simply what overflow looks like from the outside.
VIII. The Furthest Consequence
Follow the direction of absorption to its end, and it arrives at the absorbing barrier — the furthest consequence of the flow.
A system is ergodic when the fate of one participant over time matches the average of all participants at an instant. The two part ways the moment an absorbing state exists — because once a single path reaches ruin, it cannot return to the average, and the ensemble average begins to lie. This is the insight Peters and Taleb pressed: what matters is not the expectation, but the survival of the path.
Now read the three zones through this lens. When the B absorbs the A and the reversal never comes, the system drifts toward the barrier — zero is reached, and the path is absorbed into ruin or into the living death of zombification, surviving only by raising fresh capital to feed a B that never pays. When a good toll funds a value-destroying B, the barrier is crossed not by the company but by the segment: the whole drifts down in value without reaching zero. And when the A absorbs the B, the flow runs away from the barrier — there is no path to zero in the dynamics, because the toll overflows the growth at every turn.
This is why some businesses — and not only Freesurfers — cannot easily be ruined. Any business on the right side of the flow, where the A overflows the B, is structurally distant from the barrier. Not by luck, but by the direction of its capital. And here the deepest symmetry appears: the same overflow that carries the business away from ruin is the overflow that returns capital to its owners. Safety and return are not two properties. They are one property — the direction of absorption — read at two different distances.
IX. The One Variable
We began with the toll that overflowed — the beautiful problem of a business too profitable to absorb its own success. Its mirror is the business whose growth absorbs everything it earns, and drifts toward zero. Between them lies the business that erodes a part of itself, funding the wrong growth with the right toll. Three zones, one spectrum, one variable.
The recycling spread is overflow made visible. The dual duration is the dial that reads the direction. The negative risk premium is the distance from the barrier, expressed as a discount rate. The built-in convexity is what overflow feels like to an owner. And value creation or destruction — the question of whether growth pays — is a single reading on the same dial. They are not separate truths. They are one truth, seen from different rooms.
Ask of any business the one question beneath all the others. Not how fast it grows. Not how wide its moat. Which way does the capital flow — does the toll absorb the growth, or does the growth absorb the toll? Everything else is a consequence.
Watch the Video
This content is educational and reflects a personal analytical framework. It does not constitute investment advice.