Home › Articles

Returns Come From Beta, Not Distributions: A Strategy Playbook for XQQI, QQQI and TMGN

2026-10-21 · Research note · Not investment advice; figures are as of the sources and dates cited in the text · 中文
Contents9 sections
  1. The structural differences are bigger than the names suggest
  2. Measured performance: what QQQI buys is a smaller drawdown, not a higher return
  3. Return decomposition: the VRP is real, but small, and it can reverse
  4. Regime matrix: who beats QQQ in which market
  5. Actionable strategies: rules, return sources and failure modes
  6. Bear cases, data conflicts and falsification thresholds
  7. Conclusion
  8. Further reading
  9. Frequently asked questions

Bottom line first: most of the return of these three ETFs comes from equity beta. What the option seller earns is the volatility risk premium (VRP), roughly 2% a year over the long run — a small share relative to beta — and in bull-market years it usually is not enough to offset the upside given away plus fees. Since inception QQQI has an annualized NAV total return of 19.91% versus 23.63% for the Nasdaq-100, trailing by 3.7 percentage points a year. Over the same period, however, the ATM buy-write index BXN returned only 15.36%, which shows that QQQI's out-of-the-money / call-spread design gives away about 4.6 percentage points less per year in a bull market (NEOS QQQI). XQQI has notional exposure of about 1.5x, yet its 8-month cumulative return is only +16.26%, short of even 1x NDX at +18.67%; the leverage was eaten by the call cap and the 0.98% fee (NEOS XQQI). TMGN returned +18.3% including distributions over 3 months versus +5.9% for QQQ, but that excess comes almost entirely from stock selection in an "equal-weight Mag-10" basket and high beta (AMD, PLTR), and has little to do with 0DTE premium (StockAnalysis; beta and other statistics are our own calculations). The division of labor among the three is as follows. QQQI is a "lower-volatility income version" of Nasdaq exposure, with a structural edge in sideways markets and when high IV falls back. XQQI is really a 1.15–1.5x Nasdaq position with a short put; it is not an income vehicle and only makes sense as a capital-efficient beta substitute. TMGN is a concentrated tactical Mag-10 position — small, with a short track record — suitable only for a small allocation. The three have distribution rates of 14–21%, but their 30-day SEC yields are all near 0 or negative (QQQI −0.06%, XQQI −0.33%). In other words, what the distributions hand back is mostly the investor's own total return, not extra income. All samples are very short: TMGN about 3 months, XQQI about 8 months, QQQI about 32 months since 2024-01, but with only one full calendar year (2025), and mostly in a bull market. Every conclusion below should be treated as a hypothesis awaiting falsification.

This is a research note, not investment advice and not a buy or sell signal. Figures come from the issuers, SEC filings and third-party sources cited throughout, with dates as shown in each table (mostly as of 2026-09-30 or mid-2026-10); "self-calculated" statistics such as beta and capture ratios rest on very short samples. For this site's daily price and technicals on QQQI, see the QQQI ETF page (Chinese).

The structural differences are bigger than the names suggest

Item QQQI XQQI TMGN
Issuer NEOS NEOS TappAlpha (sub-adviser Tuttle Capital)
Inception 2024-01-29/30 2026-02-02 2026-07-15
AUM $15.2 billion (10/20) $405 million (10/19) $26 million–$39 million (sources differ)
Expense ratio 0.68% 0.98% 0.88% / 0.75% / 0.71% (conflicting)
Distribution rate 14.28% (monthly) 21.04% (monthly, target 19–23%) About 19–21% (annualized, only 2 payments)
30-day SEC yield −0.06% −0.33% Not published
ROC share 2024 about 94.45%; FY5/2025 about 98.86%; 2026/9 est. 97% Est. 100% to date Unknown (sister fund TSPY 2025: 88.61%)
Long position Physically holds about 107 NDX stocks Physical NDX + synthetic long (buy call / sell put) about 50% Directly holds the 10 MGTN stocks, about 10% each, rebalanced monthly
Exposure / leverage About 1.0x, beta 0.93 Notional about 1.5x; measured beta about 1.15 1.0x; measured beta vs QQQ 1.17
Options NDX index calls, near-ATM to slightly OTM, may use call spreads; inferred roughly monthly tenor NDX calls laddered OTM, partial coverage; plus short puts Daily 0DTE OTM calls (0–10%), notional coverage about 100%±5%; index or single-stock options allowed
Tax character (US accounts) §1256 (60/40), ROC deferral §1256, ROC deferral Not §1256 if single-stock options are used; character pending the 1099

Sources: NEOS QQQI, QQQI prospectus, QQQI 19a-1, 24/7 Wall St. report on the 8937, NEOS XQQI, XQQI 497K, XQQI 19a-1, TMGN SEC 485APOS, TappAlpha TMGN, TSPY tax information, P&I/Trackinsight.

XQQI's "enhancement" is the most commonly misread feature. Buying a call, selling a put, and then selling a call above — the three legs together are economically close to a short put spread: the downside is an obligation on the full notional, while the upside is still capped. So the issuer's statement that it "does not add corresponding directional risk" is consistent neither with the structure nor with the measured 1.15 beta (XQQI 497K). TMGN's 0DTE calls expire at that day's close, so after-hours earnings gaps are not covered at all. It collects premium only on intraday variance, and takes overnight risk fully in both directions.

Measured performance: what QQQI buys is a smaller drawdown, not a higher return

Window (sample noted) Fund Benchmark Note
QQQI inception to 2026-09-30, annualized (about 32 months) 19.91% NDX 23.63% / BXN 15.36% Cumulative 62.42% vs 76.23%
QQQI 6M to 9/30 (strong rally) 19.81% NDX 28.48% Upside capture about 70%
QQQI 3M to 9/30 (sideways) 1.14% NDX 0.58% Leads in a flat market
QQQI 2025 peak to trough (tariff sell-off) −20.00% QQQ −22.77% Downside capture about 88%
QQQI 1Y volatility / Sharpe (PortfoliosLab) 16.66% / 0.58 QQQ 19.82% / 0.72 JEPQ 14.78% / 0.78
XQQI inception to 9/30 (8 months) 16.26% NDX 18.67% Simple comparison at 1.5x notional about 28%
XQQI 6M / 3M / 1M to 9/30 27.26% / 0.55% / 4.38% 28.48% / 0.58% / 3.30% —
XQQI self-calculated daily data 8/7–10/9 (44 days) Beta 1.15, upside 1.20x, downside 1.10x, volatility 16.4% QQQ volatility 14.1% Single market regime, closing prices
XQQI one month to about 7/31 −7.21% QQQI −4.5% Downside amplified
TMGN self-calculated 7/28–10/21 (61 days) +18.3% incl. distributions, beta 1.17, R² 0.84, volatility 23.4% QQQ +5.9%, 18.3% Up days (n=11) 1.40x, down days (n=8) 1.02x
TMGN 10/14–10/16 sell-off −5.8% QQQ −4.6% AMD, AVGO, PLTR combined about 31%

Sources: NEOS QQQI, PortfoliosLab, NEOS XQQI, 24/7 Wall St. XQQI, StockAnalysis XQQI, StockAnalysis TMGN; self-calculated statistics are computed by our research team on very short samples.

Return decomposition: the VRP is real, but small, and it can reverse

The academic evidence is fairly consistent. Israelov and Nielsen (AQR) split a covered call into three pieces: equity beta dominates risk and return; the short-volatility piece is under 10% of total risk, with a Sharpe of about 1.0, contributing roughly 2% a year; and the remaining roughly one quarter of risk comes from "equity reversal / timing" exposure, which is almost entirely uncompensated (AQR; Quantpedia). The source of the VRP: from 1990–2024 the VIX averaged 19.59% while S&P realized volatility was 15.50%, a gap of about 4.09 volatility points; but from July to November 2008, realized volatility actually exceeded implied volatility (CFA Institute). In 1986–2011, a period without a sustained major bull market, BXM gained a cumulative +830% versus +807% for the S&P 500 (Mondo Visione). But over the last 10 years BXM had 64% upside capture and 70% downside capture — an unfavorable asymmetry (ProShares). QYLD's ATM monthly write is the most extreme counterexample: a 10-year total return of about 155%, while QQQ's price alone rose about 572% and QYLD's share price fell about 35% (24/7 Wall St., secondary source). On 0DTE, OptionMetrics backtested SPX from 2022-05 to 2026-09: selling a 25Δ call daily produced a CAGR of 16.5% and a Sharpe of 1.04, about 1.9 percentage points a year more than 30DTE. Selling only when skew was below its median raised that to 19.8% / 1.14. The conclusion: "0DTE is not special; timing is" (OptionMetrics).

Applying this framework to the three funds (inference below). QQQI: it trails NDX by 3.72 percentage points a year; net of the 0.68% fee, the options overlay cost about 3.0 percentage points in a bull market running at 23.6% annualized. That equals "VRP of about +2" minus "about 5 given away", far better than BXN's roughly 8.3 percentage points, consistent with AQR's inference that "OTM or partial coverage reduces uncompensated timing exposure". XQQI: over 8 months it fell about 12 percentage points short of 1.5x notional, and at best broke even. Its excess source is leveraged beta, not VRP. TMGN: without an MGTN index series, "basket vs QQQ" cannot be separated from "overlay vs basket". The current beta of 1.17 already explains most of the performance, and upside capture above 1 means the 0–10% OTM cap was rarely hit in the sample.

Regime matrix: who beats QQQ in which market

Market regime QQQI vs QQQ XQQI vs QQQ TMGN vs QQQ
Sharp rally Lags (6M −8.7 percentage points): near-ATM monthly calls get blown through Roughly flat or slightly behind: 1.5x offset by the cap (27.26 vs 28.48) Uncertain: basket beta helps, but single-day jumps (earnings-gap days) get capped; the 1.40x in the sample may be luck
Gradual rise Slightly behind Leads: beta about 1.15–1.2, plus premium (1M 4.38 vs 3.30) Leads: on small up days the price stays below the strike, so premium is a net gain
Sideways Leads (3M 1.14 vs 0.58), harvesting VRP Flat (0.55 vs 0.58); fees offset the premium Best in theory: collects theta daily; no evidence yet
Sharp drop Slightly better (−20.0 vs −22.8) Worst: falls 1.1–1.5x, short put has no floor Worse: beta 1.17 plus concentration (−5.8 vs −4.6)
V-shaped rebound Lags badly: monthly cap locks out the rebound Worst path: short put realizes losses at the low, rebound is then capped, distributions still paid Relatively better: the cap resets daily
High IV Premium gets richer; leads if IV falls back Premium gets richer, but usually alongside a drawdown Single-stock IV high, premium rich; overnight gaps uncovered

Actionable strategies: rules, return sources and failure modes

The percentages below are illustrative ranges for a typical experienced investor, not personalized advice; the switching rules currently have no rigorous public backtest and should be treated as hypotheses to be tested.

Strategy If (trigger) Then (action) Return source Failure mode
Core / satellite Long-term holder of Nasdaq exposure QQQ as the core; QQQI 10–30% of the Nasdaq allocation; XQQI ≤10%, replacing part of QQQ to free up cash; TMGN ≤3–5% Beta plus some VRP; lower volatility In a long bull market QQQI trails by 3–6 percentage points a year; treating XQQI as an income holding understates drawdowns
IV / regime switch VXN minus NDX 20-day realized volatility > 4 volatility points, VXN falling from its high, and NDX above its 200-day moving average Add to QQQI / TMGN Harvest VRP when it is wide (long-run average about 4 points) Volatility spikes again (2008-type reversal); mistimed
Trend first VXN < 20 and strong trend Rotate from QQQI back to QQQ Avoid regimes where premium is thin and upside is given away Whipsawed when the regime turns abruptly
Cut leverage first NDX breaks below its 200-day moving average and VXN rises Trim XQQI first, then TMGN Avoid the short put's left tail A V-shaped rebound right after cutting (costly re-entry)
Reinvest vs withdraw distributions Estimated NAV change ≈ total return − distribution rate Cap withdrawals at "long-run total return minus inflation", reinvest the rest Compounding stays in NAV If QQQI's total return is 8% and it distributes 14%, NAV falls about −6% a year, leaving about 54% after 10 years; XQQI at 10% total return and 21% distribution leaves about 31% after 10 years (inference)
Relative value (long QQQI / short QQQ; XQQI vs 1.5x QQQ; TMGN vs equal-weight Mag-10) Want to harvest VRP in isolation Most retail investors should not Net VRP in theory VRP is about 2%; 0.68–0.98% fees alone eat three-tenths to half of it, before borrow and financing costs and tax mismatches; TMGN is too small to short; samples too short to estimate the spread distribution. Using index options directly is cleaner
Things not to do Buying right before the ex-date, chasing distribution rates Don't — NAV drops by the same amount on the ex-date; ROC only lowers cost basis — the tax is deferred, not eliminated; an SEC yield ≤0 means the distribution is not income. Placing market orders in TMGN, or expecting a QQQ hedge (R² 0.84) to fully cover it, are both mistakes

Bear cases, data conflicts and falsification thresholds

The strongest bear case on QQQI: the 14% annual distribution holds up because NDX has compounded at about 23.6% a year; even so, NAV price only went from $49.14 to $54.16 (Digrin). A run of several years of 0–5% NDX returns would inevitably erode NAV; and with the top ten holdings about 47% concentrated in AI giants and downside capture of about 88%, its defense is limited. XQQI: it is a 1.5x notional position with a short put, packaged as "income". By inference, a 2022-type drawdown of about −33% would produce a loss of about −45~−50%, while still paying about $0.85 a month, creating a NAV ratchet effect; the 1.5x synthetic exposure could also be forced to deleverage at the lows under the 18f-4 VaR limit (unverified). TMGN: AUM of only about $30 million carries closure risk; the 10 equal-weighted holdings are concentrated in semiconductors, PLTR and TSLA; overnight gaps are entirely uncovered; tax character and fees are both still unsettled. Its strong showing so far rests on 11 up days.

Conflict One side Other side Assessment
TMGN expense ratio 0.88% (FAQ and all third parties) 0.75% (SEC draft), 0.71% (website fee table) Use 0.88%; defer to the final 497
TMGN holdings Directly holds 10 stocks (SEC) "Does not hold the constituents" (press release republished by Taiwan News) Direct holdings more credible; MGTN weekly calls also appear in the holdings (ETF Beacon), meaning it uses more than 0DTE
QQQI strike "Near the current price" (prospectus) "Out of the money, not fully covered" (ETF.com) Treat as near-ATM to slightly OTM with discretionary coverage; coverage ratio not published
XQQI distributions Issuer $0.856 / 0.922 / 0.877 24/7 Wall St. $0.659 / 0.622 / 0.677 Use the issuer's figures
XQQI inception date 2026-02-02 (issuer) 1/29, 2/1, 2/3 (third parties) Use the issuer's date
Indicator to monitor Falsification threshold (re-evaluate if triggered)
QQQI NAV Falls below $47 (the 2025-03 low) without a corresponding NDX drawdown
QQQI distributions Monthly distribution below $0.55, or rolling 12-month total return trailing NDX by more than 8 percentage points (currently about 6)
XQQI vs 1.5×NDX Rolling 6-month total return below 1.0×NDX (in a rising market), meaning the leverage has failed; downside capture above 1.5x
XQQI distributions / size Monthly distribution below $0.70, or AUM below $150 million
TMGN survival AUM below $25 million, shrinking volume, widening discount
TMGN behavior Rolling 60-day R² below 0.70; capture below 1.0 on days QQQ rises more than 1%, while downside capture exceeds 1.1
Tax character Early-2027 1099: TMGN's ROC share and whether §1256 applies
Premium environment VXN staying below 20 for a long time (thinner premium, pressure on all three funds' distributions)

Conclusion

The real trade-off is not "high distributions vs low distributions" but which way to hold Nasdaq beta. QQQI shows that good overlay design can cut the bull-market giveaway from roughly 8 percentage points a year at ATM to roughly 3 percentage points — but not to zero. XQQI shows that leverage plus a cap nearly cancels out in a sharp rally, yet is kept in full in a sharp drop. TMGN's excess return so far is "stock selection", not "income". Investors can exploit only two edges: overweighting QQQI in sideways markets when the VRP is wide and IV is falling back, and using §1256 and ROC deferral in US taxable accounts. Everything else is beta, and the cheapest way to hold beta is still QQQ.

For education and research only, not investment advice; all samples are short and mostly from bull markets.

Further reading

Frequently asked questions

Is QQQI worth buying?

It depends on what you want. In the roughly 32-month sample through 2026-09-30, QQQI's annualized NAV total return was 19.91%, behind the Nasdaq-100's 23.63%, but in the 2025 tariff sell-off it fell −20.00%, less than QQQ's −22.77%; and in the 3-month sideways stretch its 1.14% also beat NDX's 0.58%. If you want lower volatility and to collect premium in flat markets, QQQI has a structural edge; if you want the highest return in a long bull market, QQQ is still better. The sample is short and mostly from a bull market; not investment advice.

Why does QQQI rise less than QQQ?

Because the calls it sells give away the upside. QQQI trails NDX by about 3.72 percentage points a year; net of the 0.68% fee, the options overlay costs about 3.0 percentage points in a bull market. In the strong 6-month rally (to 9/30) QQQI returned 19.81% versus 28.48% for NDX, an upside capture of about 70%. Still, the ATM buy-write index BXN returned only 15.36% over the same period, so QQQI's OTM / spread design already gives away much less.

Are the high distributions of QQQI, XQQI and TMGN real income?

Mostly not. The three have distribution rates of 14–21%, but their 30-day SEC yields are all near 0 or negative (QQQI −0.06%, XQQI −0.33%); about 94–99% of QQQI's distributions are classified as return of capital (ROC). What the distributions hand back is mostly the investor's own total return; NAV drops by the same amount on the ex-date, and ROC only defers tax rather than eliminating it.

How is XQQI different from QQQI?

XQQI has notional exposure of about 1.5x: besides selling calls, it builds a synthetic long by buying calls and selling puts, which is economically close to a short put spread — it bears the full notional loss on the way down while upside is still capped. Its measured beta is about 1.15 and its expense ratio is 0.98% (QQQI: 0.68%). In 8 months since inception it returned a cumulative +16.26%, short of even 1x NDX at +18.67%. It is more a capital-efficient leveraged beta substitute than an income vehicle.

Why has TMGN beaten QQQ recently?

Mainly stock selection, not premium. TMGN directly holds 10 equal-weighted large-cap tech stocks; over 3 months it returned +18.3% including distributions versus +5.9% for QQQ, with a beta to QQQ of about 1.17, and high-beta holdings such as AMD and PLTR contributed the most. It sells 0DTE OTM calls daily, so after-hours earnings gaps are not covered; it has only about $30 million in assets and about 3 months of data, and in the 10/14–10/16 sell-off it fell −5.8%, deeper than QQQ's −4.6%.

What happens if I withdraw all the high distributions over the long run?

NAV gets eroded. Roughly, NAV change ≈ total return − distribution rate: if QQQI's total return is 8% and it distributes 14%, NAV falls about −6% a year, leaving about 54% after 10 years; XQQI at 10% total return and 21% distribution leaves about 31% after 10 years (inference). Cap withdrawals at "long-run total return minus inflation" and reinvest the rest, so that compounding stays in NAV.

More articles

← All articles Open today's market dashboard →

Contents