Stock

Portfolio Rules

The 24 published rules the IPOSight model portfolio runs on — entry, exit, sizing, execution and falsification — all written before the first position was opened.

These are the rules the model portfolio runs on. All of them were written before the first position was opened, and they are published here so that the record can be judged against the method rather than the other way round. A rule set that can be revised after the fact proves nothing.

What this is not. The portfolio is a disclosed rule set executed on notional capital. It is not advice, not a recommendation, and not a signal service. Nothing here tells you what to buy or sell, at what price, or in what size. If a rule below is ever changed, the change is dated and the old wording stays visible — a rule book that quietly rewrites itself is worthless.

Universe and size

1

Universe

Companies on the watchlist: listed after 2021 and still trading today.

2

How large the basket is

There is no target number of holdings. The ceiling is 200 positions, but entry is decided by the 68-point threshold — which is why few names are in at present, and why the basket grows by itself as more companies grow into the threshold. 85% of the capital works; the rest is the reserve for later additions, so that a new name never has to be funded by selling an old one.

8

Position size

The basket is sized as a whole: the raw weight comes from a company’s score relative to the others selected (0.8–1.4×), then normalised to the capital available — so the same logic works at 12 positions and at 200. No single position may exceed 8%.

9

Sector limit

No single industry may take more than 33% of the portfolio.

13

Concentration

A weight above 15% is cut back to 10%.

How a company gets in

3

Two entry tracks

The core track looks for a working business and an improving analyst picture. The dip track looks for a share price that has fallen hard while the outlook holds up — there the cause of the fall is often supply or sentiment rather than the business. Each has its own scoring, and every holding shows which track it came in on.

4

Entry filter

At least 4 analysts; market capitalisation above $500M; past the first 45 days after listing; not in distressed territory; a buy consensus, or a hold with an improving trend; 15–200% upside to the consensus target. The 45 days are not the quiet period — that is 10 days under FINRA 2241 as it now stands, and does not apply to emerging growth companies at all. The 45 days are there because before then there is usually neither analyst coverage nor a reported quarter.

5

The 200% ceiling

If the consensus target is more than three times the share price, the company is skipped. A gap that wide is not a buy signal: it means the market is in sharp disagreement with the analysts, and in that argument the market is usually right. The scoring curve peaks around 80% upside and falls away above it.

6

The dip track

At least 40% below the IPO price, with 5 or more analysts still on buy, a recommendation index that is not deteriorating, no recent downgrade, past the lock-up expiry, and at least 30% of room to the target. The scoring rewards a fall of around 60% most, but reverses below −75%, because that is where the risk of failure starts to rise. The strongest signal of all is a price that is falling while the ratings are improving.

7

Fundamentals

Where the fundamental figures are available the scoring uses them, with one and a half times the weight on the dip track, because there the whole question is whether the business is intact. One exclusion stands on its own: the value trap. A company that has fallen hard, is shrinking, and cannot cover its short-term liabilities is not bought at any score.

How a company gets out

10

Taking profit

At +150% with less than 10% of the target left: halve the position. Above +250% with the target exhausted: close it. And if a share has tripled within a year of listing and the target is used up: halve it — the fast-rally rule.

11

Broken thesis

If the consensus deteriorates to hold or sell and the recommendation index falls below −1.0: full close. On a sell majority, immediately.

12

Losses

Below −40% the position is closed only if the fundamentals are deteriorating too. If the analyst picture holds up, the position stays. Sitting through a deep fall is deliberate, not an oversight.

14

Default

Hold. There is no time-based exit; the portfolio is built for the long run.

18

Event days

A lock-up expiry on a holding does not trigger an automatic sale, but it does trigger a mandatory re-score within five days — the supply shock and the thesis are two separate questions. There are two cases of immediate full closure: a going-concern note in a filing, and a delisting notice.

21

Corporate events

In a cash acquisition the position closes at the announced price and the money returns to the reserve. In a share exchange the shares received carry the position forward at the original cost. Splits and ticker changes are bookkeeping and do not touch the thesis.

Execution and discipline

15

Execution

Nothing is traded on the day of the signal. Entries and exits are booked at the close of the first trading day after the signal. This is the one rule the whole record’s credibility rests on: without it there is no way to tell a decision made in time from one made with hindsight.

16

Review

Scoring runs on Sunday and the changes take effect at Monday’s close. There are no intraday decisions. If something happens to a holding mid-week, the next Sunday run re-scores it — except for the two immediate cases under rule 18.

17

Liquidity

A position may not exceed 5% of the twenty-day average volume. The measure is absorb days: the holding divided by average daily volume gives the number of normal trading days it would take to exit. Anything needing more than three days looks fine on paper and cannot be closed in practice.

19

News

Only a first-tier source can trigger a trade: the company’s own statement, or a regulatory filing. Two independent news agencies reporting the same thing is enough for an article but not for a position. A single source, or social media, is enough for neither.

20

Turnover

There is no monthly cap on changes — the pace is set by the market, the news and the data, not by a number. In exchange, every single move must name the trigger and the rule it falls under. The discipline does not come from a limit on the count; it comes from not being able to act without stating a reason.

22

Data freshness

If the data behind a share is more than ten days old, no new purchase is made from it, and the holding is flagged. A decision made on stale data is worse than a decision postponed.

23

Benchmark

The primary benchmark is the Renaissance IPO ETF, because it draws on the same pool: recent listings. The S&P 500 is the secondary benchmark. Both start on the same day with the same notional capital.

24

Falsification

Every position records not only why it was opened but what would disprove it. If that is not written down, the position cannot be opened. Without it, every exit would be a rationalisation after the fact.

Rules 1–14 were in force from the start of the portfolio on 31 October 2026. Rules 15–24 were added on 12 November 2026; they codify execution, review, liquidity, event handling and falsification, and none of them reverses an earlier rule. The scoring method behind the entry thresholds is set out on the Method page; the current holdings are on the Model Portfolio page.

Changes to the rules

14 November 2026 — rule 7 extended. The scoring now reads the direction of the fundamental figures alongside their level. Four signals are taken from the last five reported quarters: whether revenue rose in each of the last three, whether growth is accelerating or slowing, whether the net result of the last two quarters improved on the two before them, and whether a loss turned into a profit. Together they move a company by at most 12 points out of 100, and carry the same one-and-a-half weighting on the dip track as the rest of rule 7.

Why it was added. The figures already in use — margin, revenue growth, leverage, liquidity — describe where a business stands. They say nothing about where it is heading, and an 8% margin is good news after 2% and bad news after 15%. A company returning to profit scored no differently from one that had never left it.

What it does not change. The wording of rules 1–24 above stands unaltered; nothing was reversed or removed. The ten-factor score published in company analyses is a separate measure and is unchanged. Percentage changes are computed only from a meaningful base, so a quarter near zero cannot produce an extreme reading, and where fewer than four quarters are available the direction is scored as neutral rather than guessed.

In force from the scoring run of 16 November 2026. The portfolio opened on 31 October 2026, so the record up to that run was produced under rules 1–24 as written above. Measured on 220 companies before the change, the median effect was one point: on the core track 13 companies rose above the 68-point entry threshold and 5 fell below it; on the dip track 12 rose above and 17 fell below. The data comes from the last five quarterly reports, which is the window the source provides — a year-on-year comparison therefore exists for one quarter only, and the trend is read sequentially.