Listen to the deep-dive discussion – Harvesting Freesurfers during market panics (17:13 min)
“Growth at a reasonable price.” — Peter Lynch
What if the price is not reasonable, not free — but negative?
I. Three Generations of the Same Question
Every generation of investors has asked the same question: what should I pay for growth?
Benjamin Graham said: do not pay for growth at all. Growth is uncertain. Growth requires capital, decisions, and execution. Growth can fail. Buy the toll — the asset, the liquidation value, the tangible floor — and treat any growth as a free bonus you did not pay for. The margin of safety resides in the discount on the A. The B is suspect.
Philip Fisher said: pay for growth. Pay generously. Because a dollar retained inside the business and reinvested at 20% compounds at 20%. The friction of the dividend — the tax, the reinvestment at a lower rate — destroys value. The B is everything. Pay for it. Fisher reached for the infinite hold. But every business he could buy required capital to grow. Motorola. Texas Instruments. The B was real. The B was expensive.
Peter Lynch and Warren Buffett synthesized both into GARP — Growth At a Reasonable Price. Do not overpay for the B. Do not ignore the B. Find businesses where the growth justifies the price and the price does not overestimate the growth. The PEG ratio. The 15/15 rule. A reasonable price for a reasonable B.
Three answers to the same question. Graham: the price should be zero. Fisher: the price is worth paying. Lynch and Buffett: the price should be reasonable. All three assume that growth has a cost. That the B consumes capital. That the question “what should I pay?” has a positive answer — zero, some, or a reasonable amount.
The framework has introduced a fourth answer: the cost of the B is zero. The Freesurfer grows because the wave is external, structural, and free. Visa does not pay for the digitalization of payments. MSCI does not pay for the shift to passive investing. The toll collects. The wave pushes. The cost is zero. Growth At No Price.
This post introduces a fifth answer. The cost of the B is not zero. It is negative. The business does not merely give you growth for free. It pays you to receive it.
* * *
2. The Arithmetic of Negative Cost
Consider Visa in the spring of 2026. Free cash flow per share: approximately $19. Growth rate of FCF: 12–15% per year. Cost of reinvestment for that growth: zero. The digitalization of the world’s payments is funded by governments, banks, technology companies, and consumers. Visa collects the toll. The wave arrives uninvited and unpaid for.
If the story ended here, the cost of the B would be zero. Growth At No Price. The framework as articulated in “The Freesurfer.”
But the story does not end here. Visa simultaneously returns approximately 4% to shareholders — a buyback yield of roughly 3% plus a dividend yield of roughly 0.7%. This cash is not extracted from the growth. It is not a trade-off. The growth consumes nothing, so the entire free cash flow is surplus. And a portion of that surplus is returned to the owner every quarter. While the B delivers 12–15% growth. While the moat strengthens with every transaction. While the wave advances.
The cost of the B is not zero. It is negative. You receive the growth AND you receive cash for receiving it. The business pays you to hold the growth.
Put it in numbers. For every $100 of Visa you own, you receive approximately $4 per year in cash returns (buyback + dividend). And you receive approximately $13 in FCF growth. The cost of the $13 in growth was not $4 — the $4 is returned to you regardless. The cost of the $13 was zero. And the $4 is the payment for holding an asset whose growth costs nothing. Net cost of the B: negative $4. You were paid $4 to receive $13.
* * *
III. Why This Is Not GARP
GARP operates on the axis of price — what the market charges you for the stock. The PEG ratio divides the PE by the earnings growth rate. A PEG below 1 means the price is reasonable relative to the growth. A PEG above 2 means you are overpaying. GARP is a conversation between the investor and the market. What are you charging me? Is it fair?
The negative cost of B operates on a different axis entirely — the internal physics of the business. It is not about what the market charges. It is about what the business spends to produce the growth. And the answer is: less than nothing. The business produces growth at zero cost and returns surplus cash on top.
You can have GARP with a positive cost of B. Buffett buys Berkshire at a reasonable price, but the B costs billions in acquisitions. The price is fair. The cost is real. GARP is satisfied. The cost of B is positive.
You can have the negative cost of B at any market price. Even at PE 35, Visa’s B costs less than zero — the growth is free and the surplus is returned. The market price does not change the internal physics. The PE could be 40 and the cost of B would still be negative.
GARP asks: is the price of the stock reasonable? The framework asks: is the cost of the growth negative? The two questions are independent. And the second question is the one the spreadsheet has never been built to ask.
* * *
IV. The Spectrum Completed
Earlier posts in this series introduced a spectrum of business architectures based on the cost of growth. That spectrum is now complete.
| Architecture | Cost of B | Cash Behavior | The 15/15 | GARP Label | Examples |
| A alone | High positive. B consumed by maintenance. | Recycles. 12–18 month loop. | Broken. | Growth at full price | KO, JNJ, MCD |
| A + B | Positive. Billions per cycle. | Retained. Reinvested. | Holds. Costly. | Growth at a reasonable price (GARP) | TSM, BRK, GOOG |
| A × B (v1) | Zero. Wave is free. | Overflows. Growth AND return. | Holds. Free. | Growth at no price | Framework as of The Freesurfer |
| A × B (v2) | Negative. Growth is free AND surplus is returned. | Overflows by force. Cash returned during compounding. | Holds. Pays you. | Growth at a negative cost | Visa, MSCI, SPGI, MCO |
The gradient does not end at zero. It crosses zero and continues into negative territory. The Freesurfer does not stop at giving you growth for free. It pays you to take it.
* * *
V. The Physics of Negative Cost
In the mechanical world, the 15/15 rule describes a closed loop. The business earns 15% on equity. It retains the earnings. It reinvests them at 15%. The loop requires fuel — the retained earnings. Remove the fuel and the loop stops. This is the physics of compounding as every textbook describes it: ROE × retention rate = growth.
The Freesurfer breaks this loop. The growth does not require fuel. The wave provides the growth externally. The retained earnings are not consumed. They accumulate. And because they accumulate faster than the business can deploy them — because there is nothing to deploy them on — the surplus is returned. The buyback reduces the share count. The dividend provides income. Both happen while the growth continues.
In the mechanical world, you choose: compound or return. You cannot do both because the fuel funds one or the other. In the Freesurfer world, you get both because the fuel is not needed. The loop runs without fuel. The fuel becomes surplus. The surplus becomes your payment.
This is why the cost of B is negative. The business does not spend to grow. It grows for free. And the money it would have spent — the fuel that a mechanical business would have burned — is returned to you. You receive the growth that the fuel would have produced. And you receive the fuel itself. Both.
* * *
VI. The Compression Amplifier
The Compression Harvest, described in an earlier post, identified three floors of PE compression. At Floor 1, the market pays for the A and the B. At Floor 2, the B is free for the investor. At Floor 3, the investor is paid to own the growth.
The negative cost of B reveals that Floor 3 is not a temporary event created by market compression. It is the permanent architecture of the Freesurfer. At every PE — even at PE 35 — the cost of B is negative because the business returns cash while the growth operates for free. Floor 3 is not something you wait for. It is something you are already standing on if you own a Freesurfer.
But PE compression amplifies the negative cost. The arithmetic is simple:
At PE 35: the buyback yield is approximately 2.5%. The growth is 12–15%. Negative cost of B: the 2.5% you receive while owning 12–15% growth for free.
At PE 28: the buyback yield rises to approximately 3.5%. The same 12–15% growth. The negative cost becomes more negative — you are paid more to receive the same free growth.
At PE 22: the buyback yield rises to approximately 4.5%. The growth is unchanged. The payment for holding the growth increases further. The less the market charges for the stock, the more the business pays you to hold the growth.
This is the Compression Amplifier. The PE compression does not just make the stock cheaper. It makes the negative cost of B more negative. Each point of PE compression increases the payment you receive for holding free growth. The Compression Harvest and the negative cost of B compound together — two forces pushing in the same direction. The investor who buys during compression is not merely buying growth at a discount. He is buying a payment stream that increases as the discount deepens.
* * *
VII. What the Spreadsheet Cannot Express
The DCF model has a cell for the discount rate. The discount rate embeds the cost of risk. But there is no cell for the cost of growth — the model assumes that growth is funded by reinvestment and that reinvestment reduces the free cash flow available to shareholders. The model trades growth against cash return. More growth means less cash returned. Less growth means more cash returned. The sum is constrained.
The negative cost of B breaks this constraint. Growth does not reduce the cash returned. Growth and cash return are independent variables. Both can increase simultaneously. The model has no mechanism for this. The cell that assumes reinvestment funds growth is structurally wrong. The reinvestment is zero. The growth is 12–15%. The cash return is 4%. The model cannot reconcile these three numbers because it assumes the first determines the second and constrains the third.
This is the deepest reason the Freesurfer is permanently undervalued. The model is not wrong about one variable. It is wrong about the relationship between the variables. It assumes that growth, reinvestment, and cash return are connected. In the Freesurfer, they are disconnected. Growth is exogenous. Reinvestment is zero. Cash return is maximal. Three independent variables that the model treats as dependent.
* * *
VIII. The Evolution
The framework has evolved through this series of posts. Each step revealed something the previous step had not fully seen.
The Freesurfer said: the B is free. The toll and the wave multiply. A × B.
The Permanent Wave said: the spreadsheet cannot price a free B because the model was built for a world where growth costs capital. The Free Growth Premium is permanent.
The Compression Harvest said: when the PE compresses, the free B becomes even more valuable. The investor is paid to own growth.
This post completes the arc: the B is not merely free. It has a negative cost. The business pays you to hold the growth. The buyback and the dividend are the payment. The wave is the gift. You receive both. Always. At every price. And the payment increases as the price decreases.
Graham said: do not pay for growth. Fisher said: pay for growth. Buffett said: pay a reasonable price. The framework says: you are paid to receive it.
Five generations of the same question. Five answers on a gradient from suspicion to wonder. From the margin of safety to the Free Growth Premium to the negative cost of B. From Graham’s floor to Fisher’s horizon to Buffett’s synthesis to the framework’s climax: growth that pays you.
* * *
averagingup.com