Socialism: set conversion prices by a rule ladder, not dueling experts

Adds a Valuation chapter. It covers market price, control-size arm's-length
sales, and a formula whose industry multiples come from real acquisitions
and whose profit figures are cleaned of related-party games. Disputes stay
in court (Monongahela), but each side files a final number, costs follow
reasonableness, inflated claims draw a penalty, and workers get an
independent public valuation counsel. Conversion brackets now apply to
lifetime proceeds and look through foundations and trusts. Open problems
are updated for the anticipation discount and conversion-tax durability.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01GsSJ98117VXm9t7rBra5NX

JoshHeaps.Net/Content/Socialism/conversion.md

@@ -19,6 +19,8 @@ The first draft capped compensation at a 1% stake and called it financing. Be ho
 
 **3. A steeply progressive tax on the proceeds, traced to ultimate individuals.** This is where the moral weight lands, and it lands legally. A compensation cap on a forced sale dies in court as a taking. A tax on capital-sale proceeds is just a tax — the United States ran 91% top marginal rates in the 1950s and the sky held. The founder is paid the full number on paper; most of it, above the top brackets, comes home through a channel courts have blessed for a century.
 
+Two details keep the brackets from being dodged. The brackets apply to a person's **total** conversion proceeds, not to each year's payment, so spreading a buyout over fifteen years doesn't keep anyone in a low bracket. And money routed to a foundation, trust or relative the owner controls counts as the owner's.
+
 **4. The tax proceeds buy down conversion debt across the economy.** The revenue doesn't vanish into the general fund — it is recycled as debt relief for converting firms, funds the [retirement bridge](#the-retirement-bridge), and funds [the floor](the-floor.md). **The whales at the top of the ownership chain finance everyone else's freedom.** That is the sentence that replaces "six weeks."
 
 The result: the exiting owner walks away rich — rich in *consumption*, not in structural power over ten thousand people. Which was always the actual principle. This model never promised the founder no payday. It promised **no perpetual claim and no control**, and now it applies that rule to everyone who holds capital, including the people cashing out on day one.
@@ -34,11 +36,14 @@ The machinery survives from the first draft, repurposed:
 
 ## Valuation: settled by rule, not by litigation
 
-The first draft flagged this as a hole; full-price conversion makes it *the* hole, so it gets machinery:
+Full-price conversion makes the price the thing everyone will fight over, so it is set by a ladder of rules. The short version:
 
-- **Public firms:** market price on the lookback date. No appraisal, no argument — the market already voted.
-- **Recently traded private firms:** the last arm's-length transaction price. Conservice converts at the $5 billion TPG actually paid. There is a special justice in the owner being held to the number they chose.
-- **Everything else:** a statutory formula — a sector-standard multiple of trailing multi-year average profit. Either side may demand a formal appraisal, but the formula anchors it and the loser pays costs. The goal is that appraisal is the exception, not the economy.
+- **Public firms:** the market price around the lookback date. The market already voted.
+- **Recently sold private firms:** the price an unrelated buyer actually paid for control. Conservice converts at the $5 billion TPG actually paid. There is a special justice in the owner being held to the number they chose.
+- **Everything else:** a formula. The typical price-to-profit ratio for that industry, calculated from real sales rather than set by an official, times the firm's cleaned-up average profit.
+- **Disputes go to court, but each side has to commit to one final number first.** Whoever was unreasonable pays the costs, an owner who wildly inflates their claim pays a penalty, and the workers get an independent public lawyer. The goal is that appraisal is the exception, not the economy.
+
+The full machinery, the precedents behind it, and what it still can't handle are in [Valuation](valuation.md).
 
 ## The retirement bridge
 

JoshHeaps.Net/Content/Socialism/index.md

@@ -13,7 +13,7 @@ summary: A working model for an economy where the people who do the work own and
 
 - **Workers own their company.** Every employee gets one vote from day one. A share of the profits grows with tenure until everyone who's been there long enough splits them equally. [Worker Ownership](worker-ownership.md)
 - **Companies back each other up.** Worker-owned firms pool part of their profits into a shared fund and bank. If one firm fails, its workers are moved into others instead of being left with nothing. [The Federation](federation.md)
-- **Existing companies are bought, not seized.** Workers buy their company at full price and pay for it out of its future profits. A steeply progressive tax on the sale proceeds then funds the transition for everyone else. [The Conversion](conversion.md)
+- **Existing companies are bought, not seized.** Workers buy their company at full price and pay for it out of its future profits. A steeply progressive tax on the sale proceeds then funds the transition for everyone else. The price is set by clear rules, like the market price or what the last buyer actually paid, rather than by dueling experts. [The Conversion](conversion.md) · [Valuation](valuation.md)
 - **Investment is rented, not owned forever.** Investors still fund companies and can still earn huge returns, but their claim runs out and comes with no vote. [Capital, After](capital.md)
 - **Nobody is left behind by low-margin work.** Ownership doesn't help much when there's barely any profit to own, so a strong floor covers the rest. [Ownership and the Floor](the-floor.md)
 

JoshHeaps.Net/Content/Socialism/open-problems.md

@@ -9,7 +9,7 @@ summary: An honest list of what isn't solved yet.
 
 Honest list, updated. Several first-draft holes are patched above; here is what remains, and what's new.
 
-1. **Valuation at scale.** Now the weakest link. The formula-first machinery (market price / last transaction / statutory multiple) turns most conversions into arithmetic, but the residual appraisal fights on large, never-traded private firms are exactly the fights with the most money and the best lawyers on the other side. Formula anchoring and loser-pays shrink this; nothing eliminates it.
+1. **Valuation at scale.** *(Narrowed.)* [Valuation](valuation.md) now sets prices with a ladder of rules: market price, then a real control sale, then a formula whose ratios come from actual sales and whose profit figures are cleaned of related-party games. Disputes go to court, but each side has to file one final number, costs follow whoever was unreasonable, inflated claims draw a penalty, and the workers have an independent public lawyer. Two things remain. First, **the market already knows the bill is coming.** By the lookback date, prices reflect the expected conversion, and takings law ([*United States v. Miller*](https://www.law.cornell.edu/supremecourt/text/317/369)) normally won't let the government profit from a drop its own project caused. Correcting for it is under consideration. Second, **the truly one-of-a-kind firms** with no comparables will still end up in expert-versus-expert fights. There are far fewer of them, but they aren't easy.
 
 2. **The anticipation window.** *(Mostly patched — one true residual.)* The lookback closes almost everything: ownership and price are fixed as of the day the bill is introduced, so post-introduction transfers, restructurings, and paper losses change nothing. Value-stripping between introduction and closing — the owner paying himself a giant special dividend while still nominally in charge — is closed with a mechanism private equity already uses in its own deals: the **locked box**. Price is set at the lookback date, and any leakage after it (extraordinary dividends, above-market salary, related-party debt) is deducted straight from the purchase price. What remains is the one thing no law can pre-date: extraction and emigration *before the bill is introduced*, while the movement is visibly gaining steam. Expatriation-triggered mark-to-market tax (§877A, already in US law) catches part of that; a well-advised owner with a two-year head start still finds seams. That window is the honest residual — and it shrinks the faster the bill moves from idea to introduction.
 
@@ -25,4 +25,6 @@ Honest list, updated. Several first-draft holes are patched above; here is what
 
 8. **The 1/N problem.** *(Mostly answered.)* Peer monitoring works at 30 people; at 10,000 your extra effort raises the pool by a rounding error. The answer is culture and structure — coops really do push out free riders, and the federation model keeps units small enough for monitoring to bite while pooling risk above them.
 
+9. **Will the conversion tax hold up, and hold still?** Two risks. First, the proceeds are paid out over many years, so each payment is taxed under whichever Congress is in office, and a later Congress can cut the rates. Second, a court might see a tax aimed only at conversion proceeds as part of the taking itself, which would put the "full compensation" claim back in question. A tax that applies to every very large capital gain, not just to conversions, is on firmer ground. It's also worth being precise about the 1950s precedent: the 91% top rate applied to ordinary income, while long-term capital gains were capped at 25% from 1954 to 1967 ([Capital gains tax in the United States](https://en.wikipedia.org/wiki/Capital_gains_tax_in_the_United_States)). A 90% tax on sale proceeds would be new, not a return to the past.
+
 *Patched since the first draft:* the taking (full price + progressive tax), "nobody sells again" (conversion is mandatory and universal, not sale-triggered — there is no dodge-by-holding), the fire-sale window (lookback), risk concentration ([the federation](federation.md)), the hiring disincentive ([the tenure ramp](worker-ownership.md)), and subsidy capture on the worker floor (pay workers directly).

JoshHeaps.Net/Content/Socialism/valuation.md

@@ -0,0 +1,99 @@
+---
+title: Valuation
+section: The Transition
+order: 45
+summary: How the conversion price is set by rule instead of by dueling experts, and where courts still get the last word.
+---
+
+# Valuation
+
+[The Conversion](conversion.md) promises every owner full price. That promise is only as good as the answer to one question: *full price of what number?*
+
+For most firms, the answer is easy. For a small number of giant firms that have never been bought or sold, it is the hardest problem in the whole transition, because those are the fights with the most money and the best lawyers. This chapter explains how the price is set, why it is set by rule instead of by expert opinion, and what is still unsolved.
+
+## Why not just ask the experts?
+
+Because experts disagree by amounts that would be funny if they weren't so expensive.
+
+- **Dell (2016).** When Michael Dell took his company private, shareholders sued over the price. Two respected valuation professors, using similar methods, valued the same company at $12.68 and $28.61 per share. The judge noted the two opinions "differed by 126%, or approximately $28 billion." He then ran his own numbers, landed at $17.62, and was reversed on appeal for ignoring what the market had said. ([Potter Anderson summary](https://www.potteranderson.com/insights/cases/In-re-Appraisal-of-Dell-May-31-2016); [Harvard Law School Forum](http://corpgov.law.harvard.edu/2019/05/06/delaware-ma-appraisal-after-dcf-dell-and-aruba/))
+- **Michael Jackson's estate (2021).** The estate's appraiser valued the rights to Jackson's image and likeness at $2,105. The IRS's first demand said more than $434 million, and its trial expert said $161 million. The Tax Court settled on about $4.1 million, twelve years after he died. ([GMA CPAs](https://www.gma-cpa.com/blog/king-of-pops-estate-prevails-against-irs); [Washington Valuation](https://www.washingtonvaluation.com/post/how-much-is-michael-jackson-s-estate-worth-read-the-tax-court-s-decision))
+
+The lesson isn't that experts are dishonest. It's that the value of a business nobody is buying depends on guesses about the future, and a guess can be pushed wherever the person paying for it needs it to go. An economy-wide conversion can't run on thousands of those fights.
+
+Delaware learned the same lesson. After Dell and two similar cases, its courts started giving heavy weight to real market prices over hired forecasts, and the number of appraisal lawsuits fell from 76 in 2016 to 26 in 2018. ([Harvard Law School Forum on Corporate Governance](http://corpgov.law.harvard.edu/2019/05/06/delaware-ma-appraisal-after-dcf-dell-and-aruba/)) **Real prices beat forecasts.** The whole design below follows from that.
+
+## The price ladder
+
+Every firm is priced by the first rule on this list that applies to it. Most firms stop at the first or second step and never see a courtroom.
+
+### 1. Public companies: the market price
+
+If the company's shares trade on a stock exchange, the price is what the market says it is, averaged over the trading days leading up to the lookback date (the day the bill is introduced). Averaging over a stretch of time, instead of using a single day, means nobody can spike the price with a burst of trading.
+
+### 2. Recently sold companies: the price someone actually paid
+
+If the whole company, or a controlling stake in it, was sold to an unrelated buyer within the past five years, that sale price is the price. Conservice converts at the $5 billion TPG actually paid. There is a special justice in the owner being held to the number they chose.
+
+"Arm's-length" (a real deal between unrelated parties) has to mean it:
+
+- **Control-size only.** Selling a 3% sliver to a friend at a sky-high price doesn't set the value of the other 97%.
+- **Side deals count.** If the buyer was secretly promised their money back, the price is what they really paid, not the headline.
+- **Buyers tied to the seller don't count.** A sale to your brother-in-law's fund is not a market.
+
+### 3. Everyone else: the formula
+
+For a firm with no market price and no recent sale, the price is a formula: **typical price-to-profit ratio for that kind of business, times the firm's average profit over the past several years.**
+
+A price-to-profit ratio (economists call it a *multiple*) is just how many years of profit buyers pay for a business. The ratios differ a lot by industry. Recent market data puts trucking around 10 times yearly operating profit and software around 24 times ([Damodaran, NYU Stern](https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/vebitda.html)). So the formula's two inputs are where the gaming happens, and each needs a guard.
+
+**Guarding the ratio:**
+
+- **It comes from real sales, not from an official.** The ratio for each industry is calculated automatically from actual purchases of whole companies, which large buyers already report to antitrust regulators before a merger. Nobody sets it by hand, so nobody can be lobbied to set it.
+- **This matters because it has gone wrong before.** Taiwan built its land tax on Sun Yat-sen's idea of taxing land at a declared value. Over time the declared values became numbers set by local officials, which ran at roughly half of market value, and a tax nominally charged at 40 to 60% ended up capturing less than 20% of the gains. ([Lincoln Institute of Land Policy, 1998](https://www.lincolninst.edu/app/uploads/legacy-files/pubfiles/144_lamtsui98.pdf)) Any number officials set by hand drifts toward whatever is politically comfortable.
+- **Industries are assigned by a published list, by business line.** A conglomerate is priced piece by piece, with each division using its own industry's ratio. The list is published before anyone knows which firm it will apply to, so a trucking company can't rebrand as "logistics software" to claim a software ratio.
+
+**Guarding the profit:**
+
+- **Related-party deals are repriced to market.** If a firm has been buying from or selling to other companies its owner controls, those deals are recalculated at the prices unrelated companies would have used. The IRS already does exactly this under its transfer-pricing rule, [26 U.S.C. §482](https://www.law.cornell.edu/uscode/text/26/482).
+- **Owner pay is set to market.** An owner who pays himself a token salary and takes the real money as profit (or the reverse) gets a market salary substituted in.
+- **Cuts that fake profit come back out.** An owner who stops maintaining equipment or slashes research in the years before conversion makes profit look bigger while the business gets weaker. Skipped upkeep found after conversion is deducted from the price, extending the [locked box](open-problems.md) backward.
+
+## When someone disagrees
+
+Rules can be wrong about a particular firm. A company might hold a patent that hasn't started earning yet, or land worth far more than its profits suggest. So there has to be a way to challenge the ladder's number. The design question is how to keep that door open without letting the whole economy walk through it.
+
+### Courts keep the last word
+
+The Supreme Court held in 1893 that deciding just compensation "is a judicial, and not a legislative, question." ([*Monongahela Navigation Co. v. United States*](https://www.law.cornell.edu/supremecourt/text/148/312)) Congress can't write a formula and forbid any appeal. The model accepts that. An owner who thinks the ladder shortchanged them can go to court. The ladder's number is the starting point, and the owner has to show what specifically it gets wrong.
+
+### Each side names one number, and the costs follow the numbers
+
+Before trial, each side files one final number. After trial, whoever was unreasonable pays. This is borrowed from California's eminent-domain law, where the government has to file a final offer and the owner a final demand, and the owner recovers their legal and appraisal costs if the court finds the government's offer unreasonable and the owner's demand reasonable ([Cal. Code Civ. Proc. §1250.410](https://codes.findlaw.com/ca/code-of-civil-procedure/ccp-sect-1250-410/)). Here the rule runs both ways: an owner whose demand turns out to be wildly inflated pays the costs.
+
+Committing to a single number in advance changes behavior. In Major League Baseball's salary arbitration, each side submits one figure and the arbitrator has to pick one of them, no splitting the difference. Most cases settle before a hearing, and the bigger the gap between the two figures, the more likely they settle. ([SABR](https://sabr.org/journal/article/file-and-trial-examining-valuation-and-hearings-in-mlb-arbitration/)) When you can't count on a judge meeting you in the middle, asking for the moon stops being free.
+
+### An inflated claim costs something
+
+An owner's final demand is a sworn valuation. If the court's final number comes in far below it, the owner pays a penalty on the difference. Tax law already works this way: claim a value 150% or more of the correct figure and you owe a 20% penalty, and at 200% or more it rises to 40% ([26 U.S.C. §6662](https://www.law.cornell.edu/uscode/text/26/6662)).
+
+Owners' own past numbers are fair evidence too. Many owners have valued their businesses before, under legal liability: in estate and gift tax filings, in reports to their fund's investors, in statements to lenders. Those figures are restated to the value of the whole company (estate appraisals, for instance, usually apply discounts for owning only a minority slice, which have to be added back). Then they go in front of the court. An owner who told the IRS the firm was worth $2 billion and now tells the court it's worth $6 billion has some explaining to do.
+
+### The workers get their own lawyer
+
+In a valuation fight, the workers are the ones paying the price. But before conversion, the people who speak for "the company" are the owner's own managers. They're the least likely to argue for a lower number, especially if a retention bonus is on the table.
+
+So every contested valuation has a **public valuation counsel** on the workers' side. It is independent of management, funded from the conversion tax, and able to demand documents. Its job is an accurate price, not a low one and not a high one. That matters because the government collects more tax when the price is higher, so the tax authority can't be trusted to push back on an inflated number. Any bonus management receives for going along with a conversion counts as leakage under the [locked box](open-problems.md) and comes out of the price.
+
+## Why not an earnout?
+
+An *earnout* is a common M&A deal term: the seller gets a base price now and more later if the business performs. It looks like a perfect fit, since the conversion already pays owners out of future profits. It was considered and rejected.
+
+After the sale, the buyer controls the business and the books, and the seller can only watch. Earnouts pay out about 21 cents on the dollar and end up contested at least 28% of the time ([SRS Acquiom](https://www.srsacquiom.com/our-insights/ma-earnouts-overview/)). In a conversion, "the buyer" is the workers. They could quietly shrink the profits the owner is waiting on, or the owner's lawyers could claim they did. An earnout would also pay the old owner for gains the workers create after they own the place. It swaps one fight about the past for years of fights about the present.
+
+The payout already carries the part of an earnout that's worth having. Because the owner is paid only as profits arrive, over a fixed term, a firm that was priced too high simply pays out less than its price. The owner, not the workers, carries that risk.
+
+## What's still unsolved
+
+- **The market already knows the bill is coming.** By the lookback date, share prices will reflect the expected conversion, which pulls the "market price" down. Takings law usually refuses to let the government benefit from a drop its own project caused. This is the "scope of the project" rule from [*United States v. Miller*](https://www.law.cornell.edu/supremecourt/text/317/369) (1943). A correction for this is under consideration; see [Open Problems](open-problems.md).
+- **Truly one-of-a-kind firms.** A company with no comparable sales, an unusual mix of businesses, or value locked in things that don't yet earn profit will still end up in court, facing Dell-style battles between experts. The ladder shrinks the number of these cases. It doesn't make the last few easy.
+- **The tax only lowers the stakes if it actually lands.** At a 90% top rate, an owner who wins an extra $1 billion keeps about $100 million, which is still plenty to hire lawyers with. And it only works if the brackets apply to the owner's total conversion proceeds rather than one year's payment at a time, and if money routed through foundations and trusts counts as the controlling person's. [The Conversion](conversion.md) now says so.