Jim Simons and Renaissance Technologies: What the 13F Filings Show
Jim Simons was a mathematician who in 1978 founded what would become Renaissance Technologies, a quantitative investment firm. Its latest quarterly 13F filing, for 30 Jun 2026, lists 3,140 US-listed holdings worth $72.62 billion, and no single holding is as much as 2% of that total.
Jim Simons was the mathematician who founded the firm that became Renaissance Technologies, which his foundation says “pioneered quantitative trading”. Much of what is repeated about him and his funds cannot be checked. This page sticks to what can: his foundation’s own account of his life, and the holdings reports Renaissance files with the US Securities and Exchange Commission every quarter.
How it works
The facts of his life, from his foundation. The Simons Foundation’s announcement of 10 May 2024, read on 3 October 2026, says James Harris Simons died that day in New York City at the age of 86. It says he chaired the mathematics department at Stony Brook University, that in 1978 he founded what would become Renaissance Technologies, and that he and his wife Marilyn Simons established the Simons Foundation in 1994. It says he often described his working life in three phases: mathematician, investor and philanthropist.
The firm describes itself in one sentence. Renaissance’s home page, read the same day, calls it an investment management firm that uses mathematical and statistical methods to design and run its investment programs. It lists offices in East Setauket, on Long Island, and on Madison Avenue in New York.
The public record is Form 13F. The SEC’s Form 13F FAQ, read on 3 October 2026, says managers with investment discretion over $100 million or more in Section 13(f) securities must file, within 45 days after each calendar quarter ends, and that short positions are not reported. Renaissance files as one manager, so its 13F is a single list across everything it runs, with no split by fund.
The filing record
110 quarterly reports and 18 amendments. Renaissance Technologies LLC’s EDGAR record, CIK 0001037389, read on 3 October 2026, holds 13F-HR reports from 17 May 1999, covering the quarter ended 31 March 1999, to 13 August 2026, covering the quarter ended 30 June 2026. The 53 reports filed from August 2013 onward are read here from their machine-readable cover pages, which give a holdings count and a total value.
One unit change to watch. The SEC’s Form 13F FAQ says reports filed from 3 January 2023 must round values to the nearest dollar, where before they were rounded to the nearest thousand dollars. Renaissance’s own reports switched later.
Through the quarter ended 31 March 2024 its values are in thousands. The 31 December 2022 report lists 7,099,648 Apple shares at 922,454, which is about $130 a share if read in thousands and 13 cents if read in dollars. From 30 June 2024 they are in dollars. Read naively, the total appears to jump a thousandfold in mid-2024. Every figure on this page is converted to dollars.
A worked example
One line, three dates: Nvidia. On the 30 June 2025 list, Renaissance reported 7,422,744 Nvidia shares worth $1,172,726,130. By 31 March 2026 the line had fallen to 2,526,948 shares worth $440,696,066. On 30 June 2026 it was back to 7,091,256 shares worth $1,418,869,170, the largest line on the list.
Divide value by shares and you get the price the filing used. $1,418,869,170 / 7,091,256 = $200.09 a share at 30 June 2026, against $157.99 a year earlier. The holding’s value rose 20.99% over the year while the share count fell 4.47%, so almost all of the change in value was price, not buying.
The quarter in between tells the real story. A reader who saw only the year-apart lists would think the position barely changed. The 31 March list shows roughly two-thirds of it gone and then rebuilt within one quarter. That is the problem with reading any single 13F as a decision: it is one frame from a moving film.
Palantir shows the other direction. It was the largest line a year earlier: 13,502,046 shares worth $1,840,598,254. On 30 June 2026 the line was 7,568,370 shares worth $883,001,728, 43.95% fewer shares.
The original data
The sample. 53 quarterly 13F cover pages, 30 June 2013 to 30 June 2026, and the full holdings tables for 30 June 2025, 31 March 2026 and 30 June 2026.
Size over time. The reported value ran from $36.52 billion at 30 September 2014 to a peak of $130.13 billion at 31 December 2019. The count of holdings ran from 2,753 at 30 June 2014 to 4,173 at 30 June 2022. The latest, 30 June 2026, is 3,140 holdings worth $72.62 billion. Every quarter is in the Renaissance 13F quarters file.
Spread thin by design. On the 30 June 2026 list, the largest holding, Nvidia, is 1.95% of the reported value. The ten largest together are 12.01% and the hundred largest 43.37%. The median holding is about $4.07 million, and 831 of the 3,140 are under $1 million. Every line is a share position; the list carries no options at all.
| Rank | Holding | Value | Share of total |
|---|---|---|---|
| 1 | Nvidia | $1,418,869,170 | 1.95% |
| 2 | Meta Platforms | $1,139,611,745 | 1.57% |
| 3 | Intel | $1,022,356,867 | 1.41% |
| 4 | United Therapeutics | $925,846,052 | 1.27% |
| 5 | Palantir Technologies | $883,001,728 | 1.22% |
| 6 | Alphabet (class A) | $828,174,623 | 1.14% |
| 7 | Exelixis | $730,617,099 | 1.01% |
| 8 | VeriSign | $616,264,393 | 0.85% |
| 9 | Kinross Gold | $584,529,171 | 0.80% |
| 10 | CrowdStrike | $571,561,334 | 0.79% |
The top 25 are in the Renaissance top holdings file, with each one’s share count a year earlier.
Turnover, measured line by line. Matching the 30 June 2025 and 30 June 2026 lists by security identifier:
| Count | |
|---|---|
| Holdings on the 30 Jun 2025 list | 3,522 |
| Holdings on the 30 Jun 2026 list | 3,140 |
| On both lists | 2,311 |
| Gone a year later | 1,211 (34.38% of the earlier list) |
| New in the year | 829 (26.40% of the later list, 17.66% of its value) |
Even the survivors moved. Of the 2,311 holdings on both lists, 1,048 had more shares and 1,260 had fewer; only 3 had exactly the same count. Between the March and June 2026 lists alone, 639 holdings were dropped and 566 added. Every holding on the three lists, with its share counts, is in the matched holdings file.
What the filings cannot show
Short positions. The SEC’s FAQ says they are left out, so nothing in the filing shows whether a long line is hedged by a short elsewhere.
Which fund holds what. The list is filed by the manager as one table. Nothing in it says which Renaissance fund owns a given line, or how any fund performed.
Anything between quarter ends. The Nvidia line shows a position that nearly tripled in the three months to June after falling by about two-thirds in the nine months before. A position held for weeks inside a quarter would not appear at all.
Results. A 13F lists holdings and their quarter-end value. It reports no returns, fees or profits, and this page quotes none.
When it fails
It fails as a stock tip. Each line is under 2% of the total and a third of the list is gone within a year. Buying a line because it appears is buying one dated piece of a model you cannot see.
It fails when read late. Reports are due 45 days after the quarter ends. The 30 June 2026 list was filed on 13 August 2026, already six weeks after the date it describes.
It fails when the units are missed. Comparing a 2023 total with a 2025 total without the thousands-to-dollars change produces a meaningless thousandfold jump.
And it fails as a measure of the man. Jim Simons died in May 2024. The filings after that date describe a firm, not a person, and the firm’s own description is about methods, not any one individual.
Related
A hedge fund page covers what these firms are and where Form 13F fits among their disclosures. Quantitative analysis is the method family Renaissance describes on its home page. Institutional trading explains the other public records of large-fund activity, and the Michael Burry page reads a very different manager’s 13F filings, where options made up most of the reported value in the largest reports.
Read a quant fund’s 13F as a census, not a stock tip. With more than 3,000 lines and a third of them gone within a year, any single line is a small, dated piece of a model you cannot see.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.