How Congress Replaced Cliffs with Ramps

As monetary policy becomes less predictable, statutory clocks become relatively more valuable long term planning tools.

On July 29th the Federal Reserve held its policy rate for a fifth consecutive meeting, three of its Reserve Bank presidents dissented in favor of a hike, and Chairman Kevin Warsh once again declined to tell anyone what the committee expects to do next. He has been declining since he took office in May. When a reporter pressed him at the press conference, he said market participants are learning to play the ball rather than the referee.

I want to spend a little time on why that matters, because on the surface it sounds like a communications preference rather than an economic event.

For most of the past fifteen years the Fed published a projected path for its own policy rate, a practice generally called forward guidance. That projection was never a promise and it was frequently wrong, but it was a number everyone could plan against. A company deciding whether to finance a plant, a bank pricing a five-year loan, a family office deciding whether to sell a building this year or next, all of them were doing arithmetic with the Fed’s expected path sitting somewhere inside it. The guidance worked as a public good. It was produced at the Fed’s expense and consumed by everyone else for free.

That input has now been withdrawn on purpose. The committee still meets, still votes, and still publishes a statement. What it no longer does is tell you where it thinks the rate is going. It is also doing this at a moment when its own members visibly disagree, which is what the three dissents show. Three votes for a hike against a hold is the largest bloc in one direction since 2016. So a planner today faces an unpublished path and public evidence that the people setting it have not settled the question among themselves.

None of that stops anyone from investing. Capital keeps moving. What it changes is which schedules a long-horizon decision can lean on, and that is the thread I want to follow.

The first schedule

The bond market publishes a price for time. Four weeks costs 3.70%. Thirty years costs 5.20%. That is the schedule, updated every trading day, and everything with a horizon gets priced against some point on it.

What matters in this chart is the change in shape on the day of the decision, more than the level. Short yields fell and long yields rose, and if you follow the two lines from left to right they cross somewhere between the two-year and the five-year. The curve rotated around a point in the middle rather than moving up or down as a whole.

That is a slightly odd response to a hawkish meeting. The textbook reaction to a committee leaning toward a hike is for short yields to rise, because the front end mostly expresses what the policy rate will do over the next year or two. Here the front end fell and the long end rose by more than the front end fell. Whatever the market took from that meeting, it was some reassurance about the next few quarters paired with more doubt about the next few decades.

The level says the same thing more plainly. The thirty-year sits about 157 basis points above the effective funds rate. That gap is what the market charges to lend over a long horizon rather than a short one, and it is wide. Lenders price uncertainty by demanding more yield the further out they commit, so the steeper the curve, the more they are demanding.

Two days around one meeting is thin evidence and the move may retrace. It is worth showing anyway, because it is the clearest recent picture of what long-horizon commitment currently costs.

Treasury has responded the way any borrower does when the future is murky and money is expensive but still needed, because spending does not stop. It has shortened up. Bill supply was 21.7% of outstanding marketable debt at the end of April, above the 15% to 20% band the Treasury Borrowing Advisory Committee has recommended for years. Four-week auctions now average around $101 billion against roughly $47 billion a decade ago. The longer-dated coupon auctions still run on their regular calendar, and the committee’s own framing is that bills act as a shock absorber so long issuance can stay predictable.

The government is doing more of its borrowing at the short end while it waits for other variables to resolve. It can also do so because the buyers are there. Money market funds hold roughly $7.5 trillion and grew by close to $1 trillion in a year, and stablecoin issuers and banks are competing for the same paper. The front end has deep and growing demand. The long end has to be paid for.

None of this tells an owner what to do. Consider the standoff in commercial real estate. A buyer cannot borrow cheaply enough to pay the price a seller wants, and the seller will not mark down to meet financing that has gotten expensive. Neither side is being unreasonable. They are pricing off different rates, and until those converge the transaction does not happen and the asset sits.

Nothing in the tax code fixes that. A statutory clock does not create a buyer or lower a cap rate. What it does is change the number on one side of the seller’s decision, because a seller choosing between holding and selling is comparing after-tax proceeds against the cost of staying put. That is the term Congress moved.

Which is the point of this piece. There is now a second schedule that prices time, it is written into the tax code, and at some tenors it pays more than the first one.

The second schedule

A concentrated holder has never lacked ways out. Section 1031 has deferred gain on qualifying real property since 1921. Installment sales have spread recognition across years for generations. Borrowing against an appreciated position has never been a realization event, because loan proceeds are not income. If the problem was needing cash without a taxable sale, the code has had answers for a century.

The part that was rigid was the exclusions. Those were cliffs. You either held long enough for the whole benefit or held slightly less long and got none of it.

Section 1202 is the cleanest illustration. The Qualified Small Business Stock exclusion has always been built to reward patient ownership of small companies, and under the old rule it did that in one step. Sell at four years and eleven months and you excluded nothing. Sell a month later and you excluded the entire eligible gain. The economic case for waiting accumulated continuously across those five years while the tax benefit arrived on a single day, which made the founder’s decision effectively binary.

Stock issued after July 4, 2025 works on a schedule instead. Fifty percent of eligible gain excluded at three years, seventy-five at four, one hundred at five. The cap is $15 million per issuer or ten times basis, and the company must have had gross assets under $75 million when the stock was issued, so this is a genuinely small-company provision.

That is a term structure. Three tenors, three prices, published in statute. It can be compared directly to the one the Treasury market publishes, and as far as I can tell almost nobody does the comparison.

Take a ten million dollar eligible gain and a 23.8% federal rate, the twenty percent long-term rate plus the net investment income tax. Sell at year three with half excluded and you pay $1.19 million, keeping $8.81 million. Sell at year four with three quarters excluded and you pay $595,000, keeping $9.41 million. Sell at year five and you keep all ten.

Now read those as returns rather than tax outcomes. Waiting from year three to year four turns $8.81 million into $9.41 million, which is 6.75% for the year. Year four to year five is 6.33%. Across both years it annualizes to roughly 6.5%.

The two-year Treasury pays 4.22%.

So for that gain, at that tenor, the statutory schedule is paying about 230 basis points more than the bond market for the same two years of waiting. That is the answer to the question the first half of this piece raises, which is why anyone would commit capital for years while the market is charging so much for duration. Inside the cap, the statute is currently paying more than the market is charging.

Four things keep that from being as good as it sounds, and they matter more than the headline number.

The Treasury yield is risk-free and this one is not. Every basis point of it assumes the position holds its value, and a ten percent drawdown over those two years erases the entire benefit and more.

It is an opportunity cost rather than a trade. You cannot fund it, short the other side, or sell it to anyone.

It only runs on gain inside the fifteen million per-issuer cap, and only on stock from a company that was under seventy-five million in gross assets at issuance.

And there is a state layer. California does not conform to Section 1202, so a California resident pays full state tax on that gain regardless of the federal exclusion. Every number above is federal-only.

Section 1045 reaches a similar place from another angle. Qualified stock held at least six months can be sold and the proceeds put into another qualifying company within sixty days, and when that happens the gain defers while the original holding period carries forward toward those tiers. The practical effect is that the holding period attaches to the capital rather than to any one investment. In most financial instruments time is a property of the contract, so a five-year loan contains five years and if you sell the loan the term goes with the paper. Section 1045 detaches the clock from the instrument and leaves it with the investor.

Opportunity Zones apply the same idea at larger scale, and they contain the sharpest exception to it.

The 2025 law made the program permanent and gave each rolled gain its own clock. Roll a capital gain from any asset class into a qualified fund within 180 days and that gain gets its own five-year deferral. Ten percent of it is forgiven at five years, thirty percent for a qualifying rural fund. Everything the fund earns is excluded after ten years, and at thirty years basis steps up to market automatically with no sale at all. New maps are designated each decade, the first effective January 1, 2027. Disposing of a large position across five separate years means running five independent deferrals rather than one.

The exception is December 31st of this year. Gains deferred under the original rules get recognized on that date whether or not anything is sold, and it is the same date for every holder in the country at once. Meanwhile governors have until September 28th to submit nominations for the new map, extended to October 28th, so nobody yet knows which tracts qualify in January. A gain realized this year and rolled under the old rules buys a deferral worth close to nothing, since recognition arrives at year end regardless.

That deserves more attention than it gets, because it cuts against the pattern rather than with it. The 1202 tiers and the sixty-day rollover genuinely distribute decisions, since every holder’s clock starts on their own issuance date and no two people share a deadline. The seam does the opposite. One deadline that everyone hits simultaneously concentrates behavior instead of spreading it. Both effects sit inside the same statute, and anyone arguing these provisions smooth out the timing of realizations has to account for the one that does not.

Two older provisions were left alone, and that tells you something about what Congress was aiming at. A concentrated position contributed to a Section 721(b) exchange fund can be redeemed as a diversified basket after seven years, carrying original basis, with no sale recognized, as long as the fund keeps at least twenty percent of its assets in qualifying illiquid holdings. And borrowing against a position still produces cash while realization waits. Both solve the problem of what to do once you already have a liquid concentrated position. Congress touched neither, and spent its effort on the years leading up to an exit rather than the years after it.

None of this touches the realization doctrine. Selling still recognizes gain. Borrowing still is not a sale. What changed is that the years in between now have prices on them.

Who buys when the schedule says sell

A holder reducing a position over several years needs a market that can absorb repeated selling. Two changes in the same window worked on that side, and neither had anything to do with the tax code.

The first was research. Since 2003 the Global Research Analyst Settlement had required structural separation between equity research and investment banking at the twelve firms it covered. On December 5, 2025 the SEC consented to terminating the remaining undertakings, on motions those same twelve firms had filed in June and December of that year. Coverage now falls under FINRA Rule 2241, which applies to every broker-dealer rather than to twelve of them. Commissioner Uyeda’s stated reasoning was that the older framework had suppressed research on small and mid-sized issuers. The decision was contested. Arthur Levitt, who chaired the SEC when the original conduct occurred, published an objection in the Wall Street Journal sixteen days later, and I am not going to settle who is right here. The mechanism that matters is simpler than the dispute. Research does not create liquidity, it lowers the cost of forming a view, and a wider base of institutions willing to own a stock is a deeper pool to sell into.

The second was index eligibility. Effective May 1, 2026, a new listing ranking within the forty largest Nasdaq-100 constituents becomes eligible after fifteen trading days rather than roughly three months, on five days’ notice, and the minimum public float requirement was eliminated so low-float companies take a reduced weighting instead of exclusion. FTSE Russell went to five days. S&P Dow Jones looked at the same question and said no, keeping its earnings screen and its twelve-month wait and stating that exceptions should not be granted on market capitalization alone. That last fact is the most useful one here, because three providers facing identical commercial pressure reached different conclusions, which tells you these were judgment calls rather than something the market forced.

Congress moves these dates, and it moved this one: Section 1202 was amended repeatedly between 2009 and 2015, lapsed once, and was restored retroactively before sitting untouched for a decade. When Congress does revise these provisions it tends to grandfather, so the exposure sits on clocks you have not started rather than clocks already running. Second, the comparison at the center of this piece is checkable and it can break. The 6.5% is fixed by statute and the two-year Treasury is not, so if the two-year moves above roughly 6.5% the premium disappears and the argument inverts. That is one FRED series, any morning. Third and most important, the 6.5% is a real number and it is also the smallest number in the decision. A position that falls fifty percent has fallen fifty percent whether the gain was excluded, deferred, or paid in full. The IRS takes a capped share of a gain and the market can take all of it, which matters more as planning gets more flexible, because every structure that makes holding cheaper also makes holding feel wiser.

One thing I do not know yet and neither does anyone else. The first maps under the permanent Opportunity Zone program have not been designated, so nothing tied to geography can be evaluated until they are.

Go back to where this started. In May the Federal Reserve stopped publishing its view of where the price of money is going, and last week it declined again while three of its own presidents voted the other way. That did not make anyone’s decisions harder in any mechanical sense. It removed a number that used to sit inside those decisions for free, and the market has been repricing the far end of the curve ever since.

The useful thing about the tax changes is not that they replace what was lost, because they do not. A holding period tells you nothing about what a building will be worth or where the ten-year will trade. What they do is change the shape of one question. A founder holding concentrated stock used to be asking when to sell, which is a question about price, and right now nobody can answer a question about price with any confidence. Part of that question is now a different one: what does each additional year of waiting actually pay. That version has an answer. It is 6.75% between year three and year four, and 6.33% between year four and year five, and it is written down in a statute rather than inferred from a curve.

I want to be careful about how much weight to put on that. It is one component of a decision, and it is the smallest one. The asset can move more in a quarter than the entire exclusion is worth, and that risk sits with the holder for every year of the wait. What the statutory schedule offers is not certainty about outcomes. It is certainty about one term in the calculation, at a moment when very few terms are certain, and there is a real danger in mistaking the one you can compute for the one that matters.

The schedule also cuts both ways, which is the part most easily missed. The 1202 tiers and the sixty-day rollover distribute decisions across holders, because every clock starts on a private date. December 31st does the opposite. Every gain deferred under the original zone rules recognizes on the same day this year, and the map that governs what comes next does not exist until nominations close on September 28th. Anyone reading this in August has two dates in front of them that are considerably nearer than any three-year tier.

The changes made between July 2025 and May 2026 did not replace the realization doctrine. They left it standing and put prices on the years in between. Markets still set the price of money. Statute now sets a price for waiting, and for the moment it is the higher of the two, which is a fact about how far the first one has drifted rather than a claim about how solid the second one is.

Sources. PL 119-21 as enacted, Secs. 70421, 70431, 70303, 70341; IRC Secs. 1202, 1045, 1400Z-2, 721(b), 453, 1031; FOMC statement and Chairman Warsh’s press conference, July 29, 2026; Federal Reserve Economic Data series DTB3, DGS2, DGS5, DGS10, DGS30, T10Y2Y, T10Y3M, DFF, pulled July 30, 2026; Treasury Presentation to TBAC, Q2 2026; SEC statement of Commissioner Mark T. Uyeda, December 5, 2025; Nasdaq NDX methodology change FAQ, May 2026; S&P Dow Jones Indices and FTSE Russell published methodology statements.

Kirandeep Kaur, Series 7, 66, NMLS, CFA Level I Candidate. Eigenstate Research, Money in Motion.