A friend asked me the obvious question last week. If markets are competitive and information moves fast, why do the same handful of trading firms keep posting the same kind of returns year after year? It is a fair question. The short answer has two parts. The first is structural: a small number of firms have built compounding advantages in research, capital, and infrastructure that genuinely outpace the rest of the industry. The second is the part most retail commentary skips entirely. Those firms also operate inside a tax code that hands a meaningful return advantage to certain types of trading and certain types of legal structures. If you do not understand both parts, the persistence of their lead looks mysterious. Once you do, it stops being mysterious.
This post walks through the firms, the persistence, and the tax structure. None of it is investment or tax advice. All of the numbers are sourced from public reporting and U.S. Treasury or IRS materials, cited inline.
The firms that keep showing up at the top
Renaissance Technologies, Medallion Fund. The most-cited number in finance for a reason. From 1988 through 2018, Medallion averaged roughly 66 percent gross and 39 percent net of its 5-and-44 fee structure, turning $100 invested in 1988 into around $398.7 million by 2018. The fund has been closed to outside investors since 1993. During the 2008 financial crisis, when the S&P 500 fell 38.5 percent, Medallion gained 98.2 percent net. (Cornell Capital breakdown, Quartr Renaissance overview)
Citadel Securities. The market-making arm of Ken Griffin's empire. Net trading revenue of approximately $9.7 billion in 2024, with $4.2 billion in net income. 2025 has run at roughly $12.2 billion in trading revenue, up 25 percent year over year, with first-quarter EBITDA margins around 58 percent. (Bloomberg coverage, QuantVPS top firms list)
Jane Street. Net trading revenue of more than $10 billion in 2023 (the fourth straight year doing so), $20.5 billion in 2024, with reporting in early 2026 putting the firm's 2025 trading haul above $40 billion. Jane Street accounted for 41 percent of bond ETF trading volume in 2024 and roughly 10 percent of all U.S. equity volume. (U.S. News coverage of Jane Street 2025 revenue, Global Trading on Jane Street ETF dominance)
XTX Markets. UK-based algorithmic market maker founded by Alex Gerko. Combined revenue across UK entities of £2.74 billion ($3.53 billion) in 2024, up from £2 billion in 2023, with profit jumping more than 50 percent. (Finance Magnates on XTX 2024 results, Bloomberg coverage of XTX earnings)
Hudson River Trading. Bloomberg reported HRT out-earned Citadel Securities in Q2 2025 on both revenue and profit lines, and the firm's headcount went from roughly 900 in 2024 to 1,150 by late 2025. (efinancialcareers HFT firm coverage)
DRW. Founded by Don Wilson, around 2,000 staff including more than 800 technologists. Industry coverage repeatedly notes DRW operates "almost like a family office" for Wilson, which is not a casual observation; the legal structure of a private trading firm versus a fund versus a family office has real downstream consequences for how returns get taxed. (youngandcalculated piece on quant empires)
Two Sigma, DE Shaw, Millennium, Citadel (the hedge fund), Jump Trading, Tower Research, Quadrature, Radix. The next ring of firms, each with billions in assets or revenue and each running variations on a similar theme. Quadrature and Radix get their own dedicated post on this site; see Inside Quadrature Capital and Radix Trading.
Why the persistence is real, not a fluke
Persistent outperformance in this group is not a "they got lucky for thirty years" story. Four advantages compound.
**Capital base. ** A market maker quoting both sides of a spread on $100 million of inventory captures more dollars per basis point than one quoting on $1 million. Once a firm reaches a certain size, it can quote tighter than smaller competitors, attracting flow that smaller firms cannot service, which generates more capital, which lets the firm quote tighter still. Citadel Securities and Jane Street both sit on this side of the loop.
Infrastructure as a moat. Jane Street builds nearly all its software in-house, including risk and trading systems. XTX has invested heavily in compute. Quadrature publishes that it runs more than 20,000 CPU cores and 2,000 GPUs. The cumulative effect is that a researcher at one of these firms can ship a strategy in days that would take an outside team months to instrument. (Risk.net Jane Street FX coverage)
Hiring access. The salary numbers do their own recruiting. Quadrature averages £2.8 million per head. Top quant interns reportedly clear $10,000 a week. The firms can be ruthless in selection because the queue at the door is long and self-sorting. That talent advantage compounds in research throughput.
Multi-strategy correlation diversification. A multi-pod platform like Citadel or Millennium runs dozens of independent strategies whose return streams are not perfectly correlated. The portfolio Sharpe is much higher than any individual strategy Sharpe. A retail trader running one or two strategies cannot reproduce this directly. The institutional version of this idea costs the platform a small slice of profits per pod, in exchange for a much smoother aggregate return curve.
Each of those four advantages is real. None of them are tax-related. Tax is the layer that sits on top.
The tax structure that quietly multiplies the edge
This is the part most retail commentary either skips or gets wrong. U.S. tax code does not treat all trading income the same way. The differences are large enough to matter every year.
Section 1256 contracts and the 60/40 rule
Section 1256 of the Internal Revenue Code applies a fixed treatment to certain instruments: regulated futures contracts, foreign currency contracts, non-equity options on broad-based indices, and dealer equity options. Gains and losses are marked to market at year end, and 60 percent are treated as long-term capital gain regardless of how long the position was actually held, with 40 percent treated as short-term. (26 U.S.C. § 1256, Cornell LII, Form 6781 instructions, IRS)
The math: at the top 2025 federal bracket, the long-term capital gain rate is 20 percent and the ordinary rate is 37 percent. A blended 60/40 rate of (0.60 × 20) + (0.40 × 37) is 26.8 percent. That is 10.2 percentage points lower than ordinary income, applied to every dollar of futures profit, regardless of holding period. (Schwab trader tax explainer, Highstrike futures 60/40 walkthrough)
The instruments that qualify are also conveniently the instruments that dominate institutional risk transfer: ES futures, Treasury futures, FX futures, broad-based index options. A market maker quoting these products is paying a structurally lower tax rate on its winning trades than an investor doing the same work in single stocks.
Section 475(f) mark-to-market election for active traders
A trader who qualifies for Trader Tax Status can elect Section 475(f) treatment. The election does two things. It marks all securities positions to market at year end, and it converts trading gains and losses from capital character to ordinary character. (IRS Topic 429, IRS Section 475 FAQ, Green Trader Tax on Section 475)
That sounds like a tax penalty, since ordinary rates are higher. The advantage is on the loss side. Capital losses are limited to $3,000 per year against ordinary income, with the rest carried forward. Ordinary trading losses under 475(f) are fully deductible against any other income with no annual cap. For a trading business that experiences a serious drawdown, the difference between using the loss now versus carrying it forward at $3,000 per year for the rest of the trader's life is enormous. (Schwab on 475 election)
The IRS recently issued Rev. Proc. 2025-23, which significantly tightens the rules for revoking a 475(f) election. Once revoked, a fund cannot make a new election for five full tax years. This is worth flagging for anyone considering the election: it is not a casual switch. (Green Trader Tax on Rev. Proc. 2025-23)
The 2025 election deadline for individuals was April 15, 2025 for the 2025 tax year. New elections require both an election statement filed by the prior year's original return due date and a Form 3115 with the current return.
Carried interest
The Tax Cuts and Jobs Act introduced Section 1061, which converts long-term capital gain treatment into short-term for partnership interests received in exchange for services in certain investment businesses, unless the underlying assets are held for more than three years. (federal-tax-planning summary, Isler CPA)
For hedge fund and private equity managers, this is a meaningful constraint. For proprietary trading firms running their own capital, it is mostly irrelevant. That difference is part of why the most successful quant operators trend toward proprietary structures over external-capital fund structures: the tax handling of incentive economics is cleaner.
The family office structure, and why it matters
A family office is a private organization that manages investments and affairs for a single wealthy family or, in a multi-family structure, several families. The key tax question for any family office is whether its activity rises to a Section 162 trade or business, which permits ordinary deduction of expenses, or whether it is treated as a Section 212 investment activity, whose miscellaneous itemized deductions were suspended through 2025 by the TCJA and have now been made permanent under recent legislation.
The pivotal case is Lender Management, LLC v. Commissioner, decided by the U.S. Tax Court on December 13, 2017. The Tax Court held that the family office of the Lender's Bagels family was carrying on a trade or business as an investment manager, not operating as a passive investor. Lender Management was therefore entitled to deduct its operating expenses under Section 162. The Court emphasized that the office served multiple branches of the extended family with distinct goals and risk tolerances, employed full-time staff, used formal investment processes, and earned compensation tied to performance. (Anchin analysis of Lender ruling, BDO on Lender structuring, ACTEC Foundation, The Lender Loophole)
The downstream effect of that ruling, combined with TCJA's suspension of Section 212 deductions and the more recent permanent extension under the One Big Beautiful Bill Act, is that family offices have a strong tax incentive to structure themselves so they meet the Lender criteria. That generally means:
- A separate legal entity, typically an LLC, that contracts with multiple family branches.
- Documented services that go beyond personal administration: investment research, asset allocation decisions, financial planning.
- Compensation that ties some portion of pay to investment outcomes, mirroring how an outside adviser would be paid.
- Substantive staffing rather than a desk in the founder's primary office.
- Family branches with sufficiently distinct interests that the office is genuinely managing for multiple clients.
When the structure works, the office can deduct operating expenses, retain the long-term capital gain treatment of underlying portfolio assets, and apply Section 1256 60/40 treatment on the futures portion of its book. That combination is one of the reasons private trading firms like DRW are sometimes described as "operating almost like family offices for the founder." The legal form pulls together several favorable tax treatments at once.
The retail-accessible analogues
You are probably not running a family office. There are, however, retail-accessible pieces of the same tax map.
Section 1256 instruments are not gated. Anyone with an approved futures account can trade ES, NQ, ZN, ZB, ZF, GC, CL, /6E, and other regulated futures contracts. Broad-based index options on SPX qualify as well. The 60/40 treatment applies to retail traders just as it does to institutions. If you are going to take the same risk you would have taken via SPY options, doing it through SPX options changes the tax outcome at year end. (Pro Trader Dashboard explainer)
Section 475(f) is available to qualifying traders. Trader Tax Status is its own qualification process. Bright lines do not exist; the IRS evaluates time spent, frequency, and intent. The most-cited summary is Green Trader Tax's, which spells out the practical bar. If you genuinely day-trade for a meaningful portion of your income, this is worth a conversation with a qualified CPA.
QSBS Section 1202 just got materially better. The One Big Beautiful Bill Act, signed July 4, 2025, expanded the qualified small business stock exclusion. For QSBS issued after July 4, 2025, the per-issuer exclusion cap increased to $15 million from $10 million, with annual inflation adjustment beginning in tax years after 2026. The aggregate gross-asset threshold for issuing companies increased from $50 million to $75 million. New tiered holding periods now allow a 50 percent exclusion at three years, 75 percent at four years, and the full 100 percent at five years. (Wilson Sonsini Section 1202 walkthrough, K&L Gates analysis of OBBBA QSBS expansion, The Tax Adviser on QSBS makeover)
For an angel investor or a startup founder, Section 1202 is one of the largest legal tax exclusions in the code. It is not a market-making advantage, but it is the same kind of structural tax benefit, available to retail-scale capital, that institutional players have been compounding inside other vehicles for decades.
Tax-advantaged accounts. An IRA or 401(k) with a brokerage option lets a retail investor compound trading gains tax-deferred (traditional) or tax-free (Roth). The friction is that the contribution caps are small relative to a serious trading account. The discipline is that the percentage saved on every taxable event compounds in exactly the same way an institutional structure does, just at a smaller scale.
What this all means for the next post
The point of cataloging these advantages is not to argue retail can match institutional desks. It cannot. The point is to be honest about which advantages are structural (capital, latency, infrastructure), which are recruiting and process (research throughput, tooling), and which are tax structure. Retail can copy a piece of the second category and a piece of the third. The first category is closed.
The companion post on this site, The Retail Quant Stack: Coinbase, ccxt, vectorbt, And A Working Backtest, shows what the second-category copy actually looks like in code: a research environment, a feature library, a backtest, and a path to live execution that respects realistic costs. None of it makes you Renaissance. All of it makes you better than the median retail trader, which is the only competitor whose seat at the table you actually have a chance of taking.
Related reading
- Inside Quadrature Capital and Radix Trading for the retail-portable lessons from two specific firms
- The Retail Quant Stack: Coinbase, ccxt, vectorbt, And A Working Backtest for the runnable companion
- A 24.8 Percent AI Crypto Portfolio, Back-Checked for the after-fees, after-tax math on a single-window claim
- What Actually Fixed My Claude Code Sessions for the research process that translates across domains
Fact-check notes and sources
- Renaissance Medallion historical performance: Cornell Capital, Quartr, Renaissance Technologies, Wikipedia
- Citadel Securities revenue and Jane Street revenue: Bloomberg coverage of 2025 trading revenue, U.S. News on Jane Street 2025 trading haul, Global Trading on Jane Street market share
- XTX Markets 2024 results: Finance Magnates, Bloomberg
- HRT and DRW context: efinancialcareers HFT coverage, youngandcalculated piece on quant empires
- Section 1256 60/40: 26 U.S.C. § 1256, Cornell LII, Form 6781 PDF, IRS, Schwab, Highstrike
- Section 475(f) Trader Tax Status: IRS Topic 429, IRS Section 475 FAQ, Green Trader Tax, Green Trader Tax on Rev. Proc. 2025-23
- Lender Management family office ruling: Anchin analysis, BDO insight, ACTEC Foundation paper, RSM US summary, Coblentz Law analysis
- TCJA impact and OBBBA permanence on Section 67(g) suspension: Sapowith Tax Advisory on family office structuring, Creative Planning on family office tax structures
- QSBS Section 1202 expansion: Wilson Sonsini, K&L Gates, The Tax Adviser, Greenberg Traurig analysis
This post is informational, not investment, tax, or legal advice. Tax law changes; the citations above reflect U.S. federal law as of mid-2025 with OBBBA changes effective July 4, 2025. State tax treatment is not addressed and varies materially. Family office and entity-level structuring decisions require an attorney and a CPA. Consult qualified professionals before structuring or trading on any of the bases described above.
