Stats

Premises1 matched1254 daily gaps2021-09-21 to2026-09-18 per selectedETF, raw/split/all.2 Worst raw gaps EEM-5.739837%,QQQ-5.357103%,IWM-5.274746%; all-5.737169%,-5.357022%,-5.272192%.3 Largest raw/all differences1.839892pp,.353658pp,.512930pp. EEM largestDec20,2023 coincides issuer .748036distribution. IWM Sep15,2026 gap raw-.448057% vsall-.188048%, issuer.750436distribution. Raw twelve-month returns25.854300/22.279661/18.925028% versus adjusted28.311639/22.869016/20.131929%; top-three relative ordering unchanged. Conclusion dividends materially affect individual gaps/return levels but not measured worst gaps here. Counterexample distributions could dominate a calmer future sample; retain both series. Selected basket descriptive, not out-of-sample edge. Source https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip adjustments raw,split,all; issuer originals in observations.

Premises:1254 aligned adjusted daily returns; sample descriptive, selected basket not independent performance evidence. Five-year correlations QQQ/IWM .778152,QQQ/EEM .692313,IWM/EEM .649814; last60 .711375,.833309,.633538. IWM fall toSMA3.72733%,EEM7.36752%,QQQ8.27186%. Annual momentum can lag a new reversal; correlation rejects assumption of three independent bets, does not change fixed allocation. Empty own trade sample cannot estimate win rate, slippage or Kelly. Source https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all

Premises: aligned adjusted daily close returns Sep21 2021–Sep18 2026,1254 observations; correlations descriptive not causal or strategy validation. Five-year QQQ/IWM .778152,QQQ/EEM .692313,IWM/EEM .649814. Latest60 .711375,.833309,.633538 respectively: geographic labels do not remove common equity risk. IWM fall to SMA only3.72733%, EEM7.36752%, QQQ8.27186%; therefore IWM has weakest regime cushion despite annual20.13193%. No intramonth discretionary exit inferred; initial eligibility refreshed before entry, afterwards fixed stops/monthly rule. Source https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all . Selected-sample/descriptive inference; no live win probability or Kelly estimate from empty ledger.

Premises:1 retained prior adverse gap g EEM.125794155019,QQQ.109666233203,IWM.138906519177;2 r=2ATR/current reference close;3 stress=N*(r+(1-r)*2g+.005)+.50, N=floorcents min(280/3,250,(30-.50)/(r+(1-r)*2g+.005));4 current equity1000. Results [{"symbol":"QQQ","size":93.33,"stop":704.90264,"target":838.105473380901,"p90":0.1616958533244174,"rr":7.049793645687354,"stress":23.108077650841185},{"symbol":"IWM","size":93.33,"stop":277.666,"target":320.5246054777956,"p90":0.12821050854556687,"rr":5.661269113738809,"stress":28.421384363225883},{"symbol":"EEM","size":93.33,"stop":65.1435,"target":74.21425092953517,"p90":0.10717963493264468,"rr":3.8082432703605487,"stress":26.413232637416957}]. Combined77.942695 includes1.50 fixed cost reserve, whereas prior76.442695 excluded that reserve; no market deterioration or size change inferred. All modeled individual losses <30. Historical sampled extremes can be exceeded, so 3% is a sizing scenario not hard guarantee; complete loss of93.33 would exceed3%. Prior3644 EEM15/429 tail exceedances and3646 bootstrap retained, not rerun as new evidence. Source https://runway.adityaperswal.workers.dev/data/stock-bars . No calibrated live win odds or Kelly from an empty execution ledger; fixed mandate equal sizing prevails.

Premises: 1 adjusted open/prior-close minus1 gaps from primary feed https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all ; 2 fixed training2021–2024 N1004 and untouched-by-this-fit holdout2025–Sep18 2026 N429; 3 current basket selected ex post so this is measurement calibration, not OOS strategy return. Results [{"symbol":"EEM","train_n":1004,"test_n":429,"train_p01":-0.023161498373743362,"train_min":-0.052591847005536096,"test_min":-0.05737169009480214,"test_exceedances":15,"test_exceedance_rate":0.03496503496503497,"test_p01":-0.03355377029926494},{"symbol":"QQQ","train_n":1004,"test_n":429,"train_p01":-0.019642697452250054,"train_min":-0.05357022225224606,"test_min":-0.04140786749482395,"test_exceedances":7,"test_exceedance_rate":0.016317016317016316,"test_p01":-0.027485668992367554},{"symbol":"IWM","train_n":1004,"test_n":429,"train_p01":-0.0225591653094287,"train_min":-0.0527219225110348,"test_min":-0.04711779448621556,"test_exceedances":5,"test_exceedance_rate":0.011655011655011656,"test_p01":-0.02526703165930107}]. Common-date thresholds exceeded in >=2 names on7 days, all3 on2. EEM lower1% training threshold has3.4965% holdout exceedance, versus QQQ1.6317/IWM1.1655%; stationarity unsupported. Historical worst equal93.33 basket overnight loss10.637725USD. Counterexample QQQ/IWM holdout minima less negative than train; no universal worsening claimed. Selected basket, adjusted-price proxy, serial dependence and provider revisions limit inference; no fees or stop implementation modeled. No new strategy proposed; retain existing larger stress envelope.

Premises:1 same-date all/split responses paired1434bars each;2 provider documents all includes cash dividends https://docs.alpaca.markets/us/reference/stockbars ;3 issuer independently lists EEM two distributions https://www.ishares.com/us/products/239637/ishares-msci-emerging-markets-etf . Source bars https://runway.adityaperswal.workers.dev/data/stock-bars . Deltas [{"symbol":"SPY","paired":1434,"different":1433,"ratio_min":0.9261785620226495,"ratio_max":1,"annual_delta_pp":1.5760769564364896,"p90_delta_pp":0.4172005003522758},{"symbol":"TLT","paired":1434,"different":1421,"ratio_min":0.829594647776466,"ratio_max":1,"annual_delta_pp":4.132788584956881,"p90_delta_pp":0.90325770973202},{"symbol":"EEM","paired":1434,"different":1367,"ratio_min":0.8851376146788991,"ratio_max":1,"annual_delta_pp":2.4573389290826197,"p90_delta_pp":0.3953725147457732},{"symbol":"EFA","paired":1434,"different":1367,"ratio_min":0.8426892950391646,"ratio_max":1,"annual_delta_pp":3.8661319021926266,"p90_delta_pp":1.035611784786218},{"symbol":"GLD","paired":1434,"different":0,"ratio_min":1,"ratio_max":1,"annual_delta_pp":0,"p90_delta_pp":0},{"symbol":"IWM","paired":1434,"different":1430,"ratio_min":0.934033446586948,"ratio_max":1,"annual_delta_pp":1.2069012127939427,"p90_delta_pp":0.3200818764750507},{"symbol":"QQQ","paired":1434,"different":1371,"ratio_min":0.966392634001973,"ratio_max":1,"annual_delta_pp":0.589354770642192,"p90_delta_pp":0.17181960145787878}]. All/split both selectEEM/QQQ/IWM; no SMA eligibility flips. IWM-SPY adjusted margin3.004713pp; SPY needs2.565342% extra cumulative gross-return correction relative to IWM to tie if other inputs fixed. This is sensitivity, not probability. EEM issuer .763266 Dec16 2025 and .350802 Jun15 2026 reconstruct28.309536% under ex-date-close reinvestment versus adjusted28.311639%, difference.002103pp. Event factor residual versus prior-close adjustment -.932622/+ .118150bp; rounding/convention can contribute, precise attribution unknown. Largest all/split annual differenceTLT4.132789pp; price-return ranking not legitimate substitute even though current selection unchanged. Counterexample: larger dividends near a future rank boundary could change selection. No sample-wide cashflow reconciliation; other ETFs/older dividends unverified independently. Validity: observed robustness supports retaining current selection only; soundness bounded by common provider error, not eliminated by two endpoints. No new strategy expectancy/backtest.

Premises:1 current basket selected retrospectively;2 last1260 aligned daily returns throughSep18;3 subset599days equalweightedbasket<0. Source https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all. Unconditional QQQ/IWM .778166,QQQ/EEM .692624,IWM/EEM .650545; downside subset .615061/.475723/.396780. Lower subset correlations are conditioned on a basket built from the same returns and restrict variation; cannot infer safe tail diversification or reduce gap envelope. Counterexample: common liquidation can impose simultaneous losses despite modest historical conditional correlation. Descriptive sample, no independent strategy expectancy or Kelly input.

Premises:1 current selected basket from3536;2 last1260 aligned close returns throughSep18;3 Pearson correlation is descriptive, not crisis conditional correlation. Correlations QQQ/IWM .778166,QQQ/EEM .692624,IWM/EEM .650545; worst single-day equal93.33 basket -15.1493USD. Source https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all. Counterexample to diversification: correlations can rise toward1 during liquidation; historical worst daily close loss neither includes stop-through-gap mechanics nor bounds next loss. Today's winners were selected retrospectively; this is not an out-of-sample strategy test. Does not justify different ranking or size. Gap scenario76.4427 remains separate.

Premises:1 adjusted daily open/priorclose gaps synchronized in NY local calendar, Sep2021-Sep2026;2 hypothetical93.33 per asset preserves common shocks;3 260 >=3calendar-day gaps versus980 one-day gaps,14 two-day holidays omitted, first calendar row omitted;4 no actual stop trigger or fills modeled. Weekend basket P01 -6.6463/worst-14.8286 versus ordinaryP01-6.0251/worst-10.2101 dollars. Weekend jointnegative31.1538% vs28.2653%. Gap correlations EEM/QQQ .716731,EEM/IWM .653395,QQQ/IWM .826848. Weekend mean-.019722 vs ordinary+.127850 dollars. 2026 descriptive slice179days worst-9.63809; NOT independent unseen holdout. Largest joint gap Aug5 2024, EEM-5.25918%,QQQ-5.35702%,IWM-5.27219%. Minute crosscheck matches5/6 daily opens; IWM193.19 versus193.15 difference.04, equivalent~.0193USD on93.33, does not change qualitative result. Daily and minute trade eligibility can differ per primary https://docs.alpaca.markets/us/docs/market-data-faq ; same provider not independent corroboration. Sound conclusion: dependence persists and weekend tails observed worse, but no established mean penalty or future bound. Counterexample: worst EEM individual gap-5.73717% occurs ordinaryMar3 2026. Selection uses today's basket retrospectively; no strategyPnL/backtest, pandemic2020 omitted, no new trade filter. Source https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all

Premises:1 refreshed top3 EEM/QQQ/IWM;2 same252 completed daily returns;3 static93.33each weights. Pairwise correlations .831269/.690365/.739377;74/252 jointnegative days. Daily static worstloss13.874044,p99loss8.254056. Historical descriptions under hindsight-selected current basket, not independent strategy validation or forward odds. Separate geographies do not imply independent bets; tail correlations can exceed observed. Source https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all . No measured live fills, winrate or Kelly calibration exists in own empty ledger.

Premises:1 technical3352 chooses three equity ETFs;2 contemporaneous SIP adjusted returns are aligned on common days;3 today's sizes are hypothetical, not historical fills. Over252sessions Pearson EEM/QQQ .831269,EEM/IWM .690365,QQQ/IWM .739377; allthree negative74/252days. Current static basket dollar daily1% quantile loss8.159653,worst13.786708. Friday hypothetical close-to-close PnL+0.424213 is NOT missed executable intraday profit. IWM Friday-.465964%,20day-4.307993%,only3.871641% aboveSMA; QQQ+.631870%,20day+1.479752%,9.017802% above; EEM+.179345%,20day+.615431%,7.953498% above. Counterexample: correlations may rise in crisis and daily closes conceal stop overshoot; selection hindsight precludes strategy win-rate/Kelly inference. No ownclosed trades so own expectancy/slippage unmeasurable. Specific mandate validation already done: no repeatbacktest, nor discretionary short-term signal veto. Source https://runway.adityaperswal.workers.dev/data/stock-bars?feed=sip&adjustment=all .

Premises:1 current EEM/QQQ/IWM sizes89.95/103.11/83.46 determine fixed dollar daily return exposure;2 aligned adjusted close returns2016–Sep17 2026;3 model decay.94 fixed before this computation, standardized loss99th percentile estimated only2016–2024 after252warmup observations. Test2025-01-02–2026-09-17,428days, training2263days. Prior-day EWMA squared return forecasts next-day scale; test data update volatility only after realization; threshold multiplier3.015193607. Four exceedances/428=.934579%, latest next-day threshold8.343642USD. This is chronological risk calibration, NOT independent OOS strategy validation: todays symbols and weights chosen with hindsight, prior runs already saw historical data, no selection/rebalancing/exits/fees modeled. No tuning sweep or gate imposed. Pair correlations EEM-QQQ/EEM-IWM/QQQ-IWM train.719423/.689657/.756580; test.772195/.682891/.802395; last60.832760/.634451/.711276. Downside189days conditional correlations.578614/.393634/.642452 subject to selection conditioning. Test actual joint daily SD3.530578 versus independence2.239607; ignoring dependence understates dispersion. Historical test worst dollar daily loss15.128603 Apr4 2025, empiricalVaR99 9.368118, ES95 7.609460. Dollar returns assume daily constant exposure for a risk lens, not buy-and-hold NAV. Valid conclusion: retain dependence-aware stress sizing, not per-lot loss interpreted as book loss. Soundness limits oneprovider, revised history, daily closes miss exit delays/tails; no confidence that99% of future losses stay below threshold. Source https://runway.adityaperswal.workers.dev/data/stock-bars . Reproducible quant.py and quant_results.json.

Premises:1 daily2016–Sep17 2026 history misses2008;2 fresh15Min SIP adjustment=all crisis2020-02-18–04-15,41sessions/82bars perasset at09:30/09:45;3 g=max(daily adverse open gap, adverse first TWO RTH15min lows relative prior close), expanding prior3115 first15low/09:45open envelope to cover first30min;4 model loss=N*[r+(1-r)*2g+.005],r=2ATR/E,50bp friction assumption. Selected lots [{"symbol":"QQQ","g":0.10966623320329438,"size":103.11,"stress":24.999049024169075,"stop":700.2816399999999,"target":832.8430715665305,"P90":16.169596547248037,"R":6.967217416051264,"normal":2.39298847793338,"gain":16.67247099986745},{"symbol":"IWM","g":0.13890651917671604,"size":83.46,"stress":24.999608792011877,"stop":278.819,"target":322.0429565001183,"P90":12.827297936488211,"R":5.5381873393008965,"normal":1.9330626072942614,"gain":10.705662857793062},{"symbol":"EEM","g":0.12579415501905977,"size":89.95,"stress":24.99829032289767,"stop":65.0035,"target":74.08335366644958,"P90":10.720899217530388,"R":3.7625773230787214,"normal":2.5629902107308324,"gain":9.643448846168583}]. Worst intraday lows allMar16 2020 09:45bar. Sizes floor cents min(cap/3,25%equity,25/model fraction); $25 budget leaves~$0.50 versus strict3%equity25.4976. Roughly equal risk-constrained notionals276.52, remainder284.43 cannot be deployed to these3 under same stress, not a signal veto. Combinedstress74.99695=8.823%equity, normalstop6.88904,targetgain37.02158 beforecosts. Daily-open-only preliminary sizes117.35/119.10/97.77 superseded. This is sizing analysis, no strategy backtest or calibrated live odds/Kelly. Counterexample larger gap/halt/delayed fill exceeds3%; lows not executable quotes. No guaranteed3% ceiling. Source https://runway.adityaperswal.workers.dev/data/stock-bars .

Premises:1 synchronized41 crisis sessions2020-02-18–04-15 and29 recent sessions2026-08-03–09-11;2 old sizes96.44/96.45/96.45 and current r=2ATR/close;3 L=sum N*[r+(1-r)*max(0,-g)+.005], stress doubles adverse g; g is previousclose-to09:30open,09:45open,or first15Min low. Full diagnostics in observations27196–98. Prior daily-open doubled stress64.6743 versus intraday09:30open69.6427 and09:45open74.1207. Old per-lot09:45stress23.0467/24.2501/26.8240, first15low20.7981/26.2136/24.7325. Largest added delay loss5.3533 onMar19 vs1.3400 recent; mean .3542 vs .0188; worsening15/41 vs11/29 days. IWM daily/intraday opening difference2.9016% confirmed1Min; same provider, not independent corroboration. Neither09:45 print nor low proves executable fill; halt can defer longer. Crisis deliberately oversamples stress; recent is descriptive comparison, not an edge holdout. Sound conclusion: daily-open stress alone insufficient for this execution constraint. No crash odds/Kelly or strategy validation inferred. Sources https://runway.adityaperswal.workers.dev/data/stock-bars and https://www.sec.gov/files/rules/sro/nyse/2020/34-88413.pdf .

Premises:1 sizes96.44/96.45/96.45;2 fresh all-adjusted2020+ open/previousclose worst gaps on2020-03-16;3 stress=N*[r+(1-r)*2g+.005], r=2ATR/close. [{"symbol":"QQQ","size":96.44,"stop":699.14315,"target":830.4731316245595,"normal":2.1229602366830784,"gain":15.593948094606814,"worst_gap":0.09460338101430432,"worst_date":"2020-03-16 04:00:00+00:00","stress":20.450581934423802},{"symbol":"IWM","size":96.45,"stop":282.74895,"target":325.93680479074,"normal":2.0502761345148643,"gain":12.368598158700099,"worst_gap":0.09305480195333693,"worst_date":"2020-03-16 04:00:00+00:00","stress":20.101221352019692},{"symbol":"EFA","size":96.45,"stop":104.93826,"target":116.97551184268325,"normal":1.5925006841611973,"gain":9.288407846549196,"worst_gap":0.11621497473587517,"worst_date":"2020-03-16 04:00:00+00:00","stress":24.1224744571582}] Correlations252 {"QQQ":{"QQQ":1,"IWM":0.7375146484818859,"EFA":0.7068026873334806},"IWM":{"QQQ":0.7375146484818859,"IWM":1,"EFA":0.7494056074136088},"EFA":{"QQQ":0.7068026873334806,"IWM":0.7494056074136088,"EFA":1}}. Prior3001 split-adjusted2016+ stress64.6655; newall-adjusted64.6743 difference.0088 from definitions, not changed market. Both each<25.5297 but joint7.60%equity. Full notional loss exceeds3%; finite scenario is not guaranteed bound. EFA retained conservative max24.1263. Gross targets37.251 vsordinary5.7657 over1–3months, payoff geometry not expected return. Source https://runway.adityaperswal.workers.dev/data/stock-bars.

Premises:1 equity850.99, fund289.34 vs221.26 previously;2 sizes96.44/96.45/96.45;3 retained split-adjusted2016–Sep11 worst gaps from prior2904 are9.457211/9.301937/11.623512%, allMar16 2020;4 loss=N*[r+(1-r)*2g+.005], r=2ATR/close. [{"symbol":"QQQ","size":96.44,"normal":2.1229602366830784,"gain":15.593948094606814,"rr":7.3453792610693736,"stress":20.44468315542464,"breach_gap":0.2430583044256341,"sma_distance_pct":8.437307785361249},{"symbol":"IWM","size":96.45,"normal":2.0502761345148643,"gain":12.368598158700099,"rr":6.032649919922496,"stress":20.094531818797652,"breach_gap":0.243614842541859,"sma_distance_pct":5.756535927075612},{"symbol":"EFA","size":96.45,"normal":1.5925006841611973,"gain":9.288407846549198,"rr":5.832592688298649,"stress":24.12629631591408,"breach_gap":0.24726510275948646,"sma_distance_pct":5.986340973057924}]. Joint64.6655=7.5989%equity, ordinary5.7657, gross all-target37.251 over1–3months, not expectation. New252dailyreturn correlations {"QQQ":{"QQQ":1,"IWM":0.7375146484818859,"EFA":0.7068026873334806},"IWM":{"QQQ":0.7375146484818859,"IWM":1,"EFA":0.7494056074136088},"EFA":{"QQQ":0.7068026873334806,"IWM":0.7494056074136088,"EFA":1}}. Larger cap +30.77% does not increase evidence. Perlot stressed losses below25.5297, but EFA margin only1.4034; gap beyond stop needed for3%loss24.31–24.73%. Full loss of96.45 exceeds3%, so no hard guarantee. Hard fund and25%position caps pass. Source https://runway.adityaperswal.workers.dev/data/stock-bars ; prior risk history retained, new adjusted correlation window.

Premises:1 current equity850.99 cap221.26 supersedes187.22;2 equal73.75/73.75/73.76;3 retain measured worst split-adjusted2016–Sep11 gaps from2904 QQQ9.457211%,IWM9.301937%,EFA11.623512%, allMar16 2020;4 scenario L=N*[r+(1-r)*2g+.005], r=current2ATR/close. {"assets":[{"symbol":"QQQ","size":73.75,"normal_loss":1.6234790279487459,"target_gain":11.925069182675784,"rr":7.3453792610693736,"double_crisis_loss":15.634543578521022,"breach_gap":0.32633587000781483},{"symbol":"IWM","size":73.75,"normal_loss":1.5677331769877785,"target_gain":9.457585424615162,"rr":6.032649919922496,"double_crisis_loss":15.365181450250295,"breach_gap":0.3268561360210779},{"symbol":"EFA","size":73.76,"normal_loss":1.2178626279287705,"target_gain":7.103296659009526,"rr":5.832592688298649,"double_crisis_loss":18.450550161606895,"breach_gap":0.33005696053958994}],"joint_stress":49.45027519037821,"normal_loss":4.409074832865295,"target_gain":28.48595126630047,"cap_increase_pct":18.181818181818187}. Joint stress5.810911%equity; gap beyond stop needed to lose3%account32.63/32.69/33.01%. No empirical scenario guarantees future loss; full loss of73.76 is8.67%account. Fixed equal-weight mandate with explicit crisis scenario retained, no claim absolute3%bound. Correlations.779–.817 from2904 invalidate independence. Gross target28.49 versus4.41 ordinary stop is geometry over1–3months, not expectation. Source https://runway.adityaperswal.workers.dev/data/stock-bars . Prior simulation2905 retained, not live win odds or Kelly; no new backtest required by fixed mandate.

Premises:1 fund187.22 versus prior144.67 is29.411765% more, equity unchanged850.99;2 proposed62.40/62.41/62.41;3 crisis-inclusive2016–Sep11 2026 split-adjusted daily opens/priorcloses N2687;4 loss scenario includes current2ATR stop then multiplicative adversegap and assumed50bp friction. Formula L=N*[r+(1-r)*g+.005], r=2ATR/close. Worst gaps QQQ9.457211%,IWM9.301937%,EFA11.623512%, allMar16 2020. Gap correlations QQQ/IWM.811520,QQQ/EFA.778810,IWM/EFA.816941. Using DOUBLE each worst gap gives13.228414/13.002589/15.611427dollars=1.554474/1.527937/1.834502%equity each below25.5297 risk threshold. Joint41.842430=4.916912% equity. Common20% extra beyondstop+50bp loses41.364711=4.860775%; common10% loses23.015787. Ordinary stops sum3.730764; all initial targets yield24.103442 gross, not expected payoff. To reach3% per lot requires39.071750/39.112107/39.405948% extra gap beyond stop under this equation. Total loss of62.41lot is7.333811%equity, so no guaranteed3%bound. An unconditional per-lot total-loss cap would need<=25.52 before fees. Conclusion: current equal lots pass explicit severe historical scenarios; diversification does not imply book3%bound. This is stress sizing, not a trade-edge backtest or probability forecast; full-loss possibility and pre2016/sample omissions remain. Sources https://runway.adityaperswal.workers.dev/data/stock-bars?timeframe=1Day&feed=sip&adjustment=split&start=2016-01-01 and user venue/account contextSep12 06:01UTC. Same vendor across all measurements, not independent corroboration. Prior chronologically separated tail check remains in2845/2846; no re-estimation needed to scale dollars.

Question: does the fixed roughly $48.22 allocation remain tolerable when fractional exits wait for RTH? Premises: (1) $850.99 equity,3%=$25.5297; (2) losses may jump through stops; (3) current selected ETFs share equity-market risk; (4) historical sample cannot bound unseen shocks. Source https://runway.adityaperswal.workers.dev/data/stock-bars?symbols=QQQ,IWM,EFA&timeframe=1Day&feed=sip&adjustment=split . Vendor SIP daily open/priorclose-1, split-only adjustments preserve ex-dividend price gaps. Five-year Sept2021–Sept2026 N1255 worst QQQ-5.3571%,IWM-5.2747%,EFA-4.5015%; QQQ/IWM worstAug5 2024, EFA Mar3 2026. Monday proxy234 observations; worst -5.3571/-5.2747/-3.5974%; not complete long-weekend classification. Pairwise gap correlations .8245/.7278/.7597; total-return daily correlations .7784/.7349/.7545. Worst joint opening loss at proposed weights $6.8617 Aug5 2024. Expanded Jan2016–Sept2026 N2687: all three worst Mar16 2020, QQQ-9.457211%,IWM-9.301937%,EFA-11.623512%; joint loss$14.651681, correlations .81152/.77881/.81694. Expansion changes which ETF has largest tail; five-year-only complacency rejected. Descriptive chronological tail check:2016–24 N2263 P1 cutoffs-2.51813/-2.55732/-2.46952%;2025–26 N424 breaches6/5/4=1.4151/1.1792/.9434%. This was not prospectively frozen; no claim of pristine holdout. Selection of today's ETFs is survivorship/selection-conditioned; history lacks2008 and deeper extremes. No probability of future bound compliance inferred. Counterexample to independence: same worst date for all3. Counterexample to historical maximum as limit: earlier2020 observations more than double EFA's five-year maximum. Conclusion: retain fixed ranks, use crisis-inclusive risk estimates and disclose common shock explicitly; no momentum re-test.