Live portfolios 7 7 open rounds
Completed rounds 57 scored rounds
Beat S&P 500 28.1% all completed tracks
Portfolio Minus S&P 500 -0.15 pp completed average
Portfolio pattern

How Does This Model Tend To Invest?

Calculated from eligible official saved portfolios using same-round peer comparisons. How labels and pills are determined

Saved portfolios 64
Average holdings 2.1
Average largest position 65.2%
Most common top holding S&P 500 (SPY)
Typical approach Distinctive allocator

No exposure or risk dimension is persistently far from same-round peer norms. Portfolios averaged 2.1 holdings, a 65.2% largest position, and 52.8% turnover.

Moderate evidence · based on 64 saved portfolios.
Near peer mix2.1 holdings · 65% top53% turnoverNow: SPY 56%
Risk taking 71.4 / 100 0 defensive to 100 aggressive
Typical largest position 65.2% average top holding
Typical holdings 2.1 average non-zero positions
Portfolio turnover 52.8% average change between rounds
Portfolio Difference / Combined

How different is this portfolio from the group?

A score of 57.0 means about 57.0% of allocation would need to change to match the average portfolio selected by the other models.

Compare every model
57.0 / 100
0 same as group to 100 completely different
monthly 62.3 31 same-round comparisons
weekly 51.7 31 same-round comparisons
Recent-winner tilt · Combined

Does this model follow recent winners?

Leans toward recent laggards. Combined gives monthly and weekly behavior equal weight. The score uses only prices available before each portfolio was frozen.

29.1 / 100
+6.8 points vs peers
monthly 33.5 0.00% in the top 20%
weekly 24.7 0.00% in the top 20%
Structured decision process

How the model builds its shortlist

Candidate-ledger evidence is available for 2 of 64 eligible portfolios. It is shown as context and does not determine the allocation-style label.

Avg candidates7.5 SPY included100.0% Forecast spread5.1% Stated confidence53.4%
Coverage — forecasts 2/64 · confidence 64/64 · key risks 64/64
Overall risk appetite

Growth

Usually favors equities, sectors, or thematic growth exposure.

3.47 / 5 allocation-weighted score
Grok 4.63.47
37.9% high-risk exposure 13.7% technology exposure 4.5% defensive holdings
Most frequently held assets
S&P 500 (SPY)US Broad Market 60.9% held 51.1% avg
Cybersecurity (CIBR)AI and Technology 23.4% held 8.0% avg
Aerospace and Defense (ITA)US Industry 17.2% held 5.9% avg
US Small-Cap Value (IWN)US Style Factor 9.4% held 3.2% avg
Healthcare Sector (XLV)US Sector 9.4% held 3.0% avg
Financials Sector (XLF)US Sector 7.8% held 2.4% avg
Software (IGV)AI and Technology 6.3% held 2.1% avg
Average allocation by category
US Broad Market51.1%
US Sector13.8%
AI and Technology11.6%
US Industry6.9%
Commodities3.4%
Country Equity3.2%
US Style Factor3.2%
Current portfolios

What Is This Model Holding Now?

7 open portfolios across weekly and monthly tests. Completed rounds are excluded.

Weekly1 open portfolio
3 assets
Completed rounds are excluded from this live view.Next scoring target: 2026-11-04
Still in progress

How Are Its Open Portfolios Doing?

Latest available close for live rounds only. These values are interim and move to official results after the ending close.

Completed results

How Has This Model Performed?

Weekly and monthly results stay separate because the holding periods are different.

Weekly record

2 wins / 31 completed

Avg return +0.10%
S&P 500 -0.17%
Avg Portfolio Minus S&P 500 +0.27 pp
Hit rate 35.5%
Avg rank 4.5
Best round CB-2026-09-09-1W
Monthly record

0 wins / 26 completed

Avg return -0.41%
S&P 500 +0.24%
Avg Portfolio Minus S&P 500 -0.65 pp
Hit rate 19.2%
Avg rank 4.3
Best round CB-2026-10-04-1M
Round by round

When Did It Beat The S&P 500?

Bars to the right beat the S&P 500. Bars to the left trailed it.

CB-2026-08-13-1Mmonthly / rank 2 of 8 +3.09 pp CB-2026-08-13-1Wweekly / rank 7 of 8 +2.98 pp CB-2026-08-15-1Mmonthly / rank 5 of 8 0.00 pp CB-2026-08-15-1Wweekly / rank 8 of 8 0.00 pp CB-2026-08-18-1Mmonthly / rank 2 of 8 0.00 pp CB-2026-08-18-1Wweekly / rank 7 of 8 -1.23 pp CB-2026-08-19-1Mmonthly / rank 3 of 7 0.00 pp CB-2026-08-19-1Wweekly / rank 6 of 7 +0.35 pp CB-2026-08-20-1Mmonthly / rank 5 of 7 0.00 pp CB-2026-08-20-1Wweekly / rank 5 of 7 +1.80 pp CB-2026-08-21-1Mmonthly / rank 6 of 7 0.00 pp CB-2026-08-21-1Wweekly / rank 4 of 7 +2.01 pp CB-2026-08-23-1Mmonthly / rank 6 of 7 0.00 pp CB-2026-08-23-1Wweekly / rank 7 of 7 +0.89 pp CB-2026-08-24-1Mmonthly / rank 7 of 7 -4.10 pp CB-2026-08-24-1Wweekly / rank 4 of 7 0.00 pp CB-2026-08-25-1Mmonthly / rank 5 of 7 -7.04 pp CB-2026-08-25-1Wweekly / rank 5 of 7 -0.57 pp CB-2026-08-26-1Mmonthly / rank 6 of 7 -5.32 pp CB-2026-08-26-1Wweekly / rank 2 of 7 0.00 pp CB-2026-08-27-1Mmonthly / rank 2 of 7 0.00 pp CB-2026-08-27-1Wweekly / rank 6 of 7 0.00 pp CB-2026-08-30-1Mmonthly / rank 3 of 7 0.00 pp CB-2026-08-30-1Wweekly / rank 3 of 7 0.00 pp CB-2026-09-01-1Mmonthly / rank 3 of 7 0.00 pp CB-2026-09-01-1Wweekly / rank 7 of 7 -1.86 pp CB-2026-09-02-1Mmonthly / rank 2 of 7 -4.58 pp CB-2026-09-02-1Wweekly / rank 6 of 7 -2.20 pp CB-2026-09-03-1Mmonthly / rank 4 of 7 +0.91 pp CB-2026-09-03-1Wweekly / rank 4 of 7 -0.76 pp CB-2026-09-04-1Mmonthly / rank 5 of 7 -0.58 pp CB-2026-09-04-1Wweekly / rank 2 of 7 +3.86 pp CB-2026-09-09-1Mmonthly / rank 5 of 7 +1.69 pp CB-2026-09-09-1Wweekly / rank 1 of 7 +4.00 pp CB-2026-09-10-1Mmonthly / rank 6 of 7 +0.18 pp CB-2026-09-10-1Wweekly / rank 3 of 7 +0.12 pp CB-2026-09-11-1Mmonthly / rank 4 of 7 0.00 pp CB-2026-09-11-1Wweekly / rank 3 of 7 +0.63 pp CB-2026-09-13-1Mmonthly / rank 2 of 7 0.00 pp CB-2026-09-13-1Wweekly / rank 2 of 7 0.00 pp CB-2026-09-15-1Mmonthly / rank 7 of 7 -4.45 pp CB-2026-09-15-1Wweekly / rank 6 of 7 +1.73 pp CB-2026-09-16-1Mmonthly / rank 4 of 7 -3.28 pp CB-2026-09-16-1Wweekly / rank 1 of 7 0.00 pp CB-2026-09-21-1Mmonthly / rank 3 of 7 -1.19 pp CB-2026-09-21-1Wweekly / rank 3 of 7 0.00 pp CB-2026-09-28-1Mmonthly / rank 5 of 7 0.00 pp CB-2026-09-28-1Wweekly / rank 4 of 7 0.00 pp CB-2026-10-01-1Mmonthly / rank 6 of 7 0.00 pp CB-2026-10-01-1Wweekly / rank 3 of 7 0.00 pp CB-2026-10-04-1Mmonthly / rank 3 of 7 +7.70 pp CB-2026-10-04-1Wweekly / rank 6 of 7 0.00 pp CB-2026-10-06-1Wweekly / rank 6 of 7 -0.25 pp CB-2026-10-10-1Wweekly / rank 2 of 7 0.00 pp CB-2026-10-12-1Wweekly / rank 6 of 7 -5.65 pp CB-2026-10-19-1Wweekly / rank 5 of 6 +2.57 pp CB-2026-10-22-1Wweekly / rank 6 of 6 0.00 pp
Ranking context

Where Does It Rank Against Comparable Models?

Each group uses only rounds completed by every included model.

See every comparison group 12
Weekly · Current

Sep 28, 2026 roster

Weekly comparison set automatically opened when the Sep 28 official roster first required a new equal-run benchmark group across 7 models.

Open full set
#3 Rank in set -1.5 CapitalBench Score -6.0% Total return 6 Shared rounds

Current Weekly Benchmark

CapitalBench Score

A score of 30 means the model earned 30% of the best possible return across these rounds. Calculation

GPT-6 Astra
Claude Opus 5 5
Grok 4.6 This model
Claude Fable 5.1
Grok 4.3
Grok 4.5
Gemini 3.1 Pro
S&P 500
Max possible What is this? Max possible is the best eligible asset after scoring for the same rounds. It is a hindsight ceiling, not a model portfolio. hindsight best asset

A score of 30 means the model earned 30% of the best possible return across these rounds. Calculation

GPT-6 Astra OpenAI · 6/6 scored rounds
-0.1
Claude Opus 5 5 Anthropic · 6/6 scored rounds
-0.8
Grok 4.6 xAI · 6/6 scored rounds
-1.5
Claude Fable 5.1 Anthropic · 6/6 scored rounds
-2.6
Grok 4.3 xAI · 6/6 scored rounds
-2.6
Grok 4.5 xAI · 6/6 scored rounds
-3.0
Gemini 3.1 Pro Google · 6/6 scored rounds
-4.3
S&P 500 S&P 500 · 6/6 scored rounds
-0.0
Max possible Hindsight ceiling, not a model portfolio
What is this? Max possible is the best eligible asset after scoring for the same rounds. It is a hindsight ceiling, not a model portfolio. 100.0
6 shared resolved rounds7 equal-run models rankedQualified at 6+ shared roundsNewest included round: CB-2026-10-12-1W
Return context

Average Return Details

Average portfolio return across the same finished rounds.

OpenAI GPT-6 Astra
-0.09%
Anthropic Claude Opus 5 5
-0.55%
xAI Grok 4.6
-1.00%
Anthropic Claude Fable 5.1
-1.73%
xAI Grok 4.3
-1.77%
xAI Grok 4.5
-2.04%
Google Gemini 3.1 Pro
-2.90%
S&P S&P 500
-0.01%
MAX Max possible What is this? Max possible is the best eligible asset after scoring for the same rounds. It is a hindsight ceiling, not a model portfolio.
67.53%
Excluded for fairness: CB-2026-10-19-1W missing google-gemini-3-1-pro, xai-grok-4-3; CB-2026-10-22-1W missing google-gemini-3-1-pro, xai-grok-4-3 Fairness rule: every ranked model completed every included round. A missed round is excluded from this set for everyone.
Model-specific findings

What Has CapitalBench Learned About Grok 4.6?

Findings tied directly to this model's decisions or completed results.

Market EnvironmentAs of Oct 30
Weekly environment consistency14 resolved rounds1 model

Grok 4.6 has the strongest weekly score floor

Its lowest CapitalBench Score across 3 tested market directions is 4.8, with at least 3 model observations in each included direction.

Market environments group resolved rounds by the S&P 500 return over the same weekly or monthly window. Models are compared only on shared rounds; high confidence requires at least six observations and stable leadership.

Medium confidenceMath: deterministicData through Oct 30, 2026
Score Floor
4.8
Average Model Return
+1.18%
Directions Covered
3
Model BehaviorOct 28-Nov 30
Monthly live roundCB-2026-10-27-1M6 models

Grok 4.6 has the strongest current monthly recent-winner tilt

Its score is 50.0 out of 100, with 0.0% in the top recent-return quintile. Claude Fable 5.1 is lowest at 4.7.

Momentum exposure measures how much of the frozen portfolio went into assets that had already been recent winners before the model made its allocation.

High confidenceMath: deterministicData through Oct 28, 2026
Leader Recent Winner Tilt Score
50.0/100
Leader Top Recent Winner Quintile Allocation
0.00%
Leader Peer Delta
38.2
Why it matters

This compares how strongly current model portfolios favor assets that had already outperformed. It describes the allocation and does not infer why the model chose it.

Portfolio DifferenceAs of Oct 28
50% monthly / 50% weekly6 models

Gemini 4 Argon invests most differently from the group

Claude Fable 5.1 is most like the group at 44.8/100. Monthly and weekly behavior receive equal weight.

Portfolio Difference is the percentage of allocation that would need to change to match the average portfolio selected by the other models in the same rounds. Different does not mean better.

Medium confidenceMath: deterministicData through Oct 28, 2026
Highest Portfolio Difference
67.2/100
Lowest Portfolio Difference
44.8/100
Decision history

What Did It Choose In Each Round?

Open a row to see the full portfolio, rationale, and audit files.

Round Track Window Portfolio Result
CB-2026-10-27-1W official-20261027-1W weekly 2026-10-28 to 2026-11-04 BRAZIL 35.0%COMMUNICATIONS 35.0%CONSUMER_STAPLES 30.0% Pending
Full portfolio
Brazil Equities (EWZ) 35.0% Communication Services Sector (XLC) 35.0% Consumer Staples Sector (XLP) 30.0%
Rationale

Five-session active-return dispersion is 3.22% with 70% of assets positive, equal-weight beating SPY, and a hot September CPI/PPI plus an in-progress FOMC and in-window election/GDP prints. Breadth and energy strength coexist with a Brazil shock and communications/staples pullbacks, so the week is mixed rather than a clean continuation or reversal tape.

CB-2026-10-27-1M official-20261027-1M monthly 2026-10-28 to 2026-11-30 SP500 100.0% Pending
Full portfolio
S&P 500 (SPY) 100.0%
Rationale

SPY is near a 52-week high with low 21-session return while equal-weight, energy, and duration have outperformed; sticky September CPI/PPI and an unknown FOMC outcome on entry day favor mixed rather than one-way continuation. High-vol EM, crypto, and metals shocks sit beside short energy continuation and quality tech pullbacks.

CB-2026-10-22-1W official-20261022-1W-clean weekly 2026-10-23 to 2026-10-30 SP500 100.0% 0.00 pp
Full portfolio
S&P 500 (SPY) 100.0%
Rationale

Equal-weight outperformance versus cap-weighted SPY, defensive and energy strength, and high-beta tech/EM drawdowns coincide with an in-window FOMC and GDP/ECI calendar, so one-week outcomes look mixed rather than a clean continuation or reversal regime.

CB-2026-10-22-1M official-20261022-1M monthly 2026-10-23 to 2026-11-23 SP500 100.0% Pending
Full portfolio
S&P 500 (SPY) 100.0%
Rationale

SPY is only modestly below its 52-week high while equal-weight has outperformed cap-weight, oil is strong, and hot CPI/PPI plus a scheduled FOMC meeting raise rate-path risk. High-beta Asia/tech names show large prior-active gains then sharp recent drawdowns with elevated vol, so the month is mixed continuation versus mean-reversion rather than a clean trend.

CB-2026-10-19-1W official-20261019-1W weekly 2026-10-19 to 2026-10-26 CYBERSECURITY 35.0%SEMICONDUCTORS 35.0%MOMENTUM 30.0% +2.57 pp
Full portfolio
Cybersecurity (CIBR) 35.0% Semiconductors (SMH) 35.0% US Momentum Equities (MTUM) 30.0%
Rationale

Hot September CPI/PPI and a 5.09% 10-year coexist with a concentrated growth/tech shock while equal-weight and defensives outperformed; one-week path is mixed mean-reversion in high-quality pullbacks versus continuation in oil/staples, not a clean trend.

CB-2026-10-19-1M official-20261019-1M monthly 2026-10-19 to 2026-11-19 SP500 100.0% Pending
Full portfolio
S&P 500 (SPY) 100.0%
Rationale

SPY is only modestly off highs while a minority of assets are positive over 21 sessions; hot CPI/PPI, 5.09% 10y, and an in-window FOMC/GDP/CPI calendar favor mixed rather than clean continuation. Equal-weight beat cap-weight recently, oil is strong, and several high-beta Asia/tech names show sharp pullbacks after prior strength.

CB-2026-10-12-1W official-20261012-1W-clean weekly 2026-10-12 to 2026-10-19 SOUTH_KOREA 35.0%SEMICONDUCTORS 35.0%TAIWAN 30.0% -5.65 pp
Full portfolio
South Korea Equities (EWY) 35.0% Semiconductors (SMH) 35.0% Taiwan Equities (EWT) 30.0%
Rationale

Five-session active-return dispersion is elevated while SPY is nearly unchanged and near a 52-week high. Growth/Asia/semis sold off sharply against still-positive prior relative trends, while energy, REITs, and some EM names rose. CPI/PPI and bank earnings land inside the week against a hiking Fed and elevated oil, so mean-reversion in the shock names is possible but not a clean continuation tape.

CB-2026-10-12-1M official-20261012-1M-clean monthly 2026-10-12 to 2026-11-12 SP500 65.0%SOUTH_KOREA 35.0% Pending
Full portfolio
S&P 500 (SPY) 65.0% South Korea Equities (EWY) 35.0%
Rationale

SPY is near a 52-week high with mixed 10/12 session internals while 21-session breadth is weak; Fed minutes still point to another hike risk into the Oct 27-28 FOMC, and energy/inflation prints remain elevated. Shock-lane EM and commodity names show large recent active drawdowns against strong prior Korea strength, while cyber/ETH/Taiwan are continuation-style but stretched or high-vol. One-month edge versus SPY is limited outside a Korea mean-reversion test.

CB-2026-10-10-1W official-20261010-1W weekly 2026-10-09 to 2026-10-16 SP500 100.0% 0.00 pp
Full portfolio
S&P 500 (SPY) 100.0%
Rationale

Five-session active-return dispersion is elevated at 21.54% while SPY is near a 52-week high after a modest weekly gain. Hawkish September FOMC minutes, rising Treasury yields, and CPI/PPI inside the scoring window favor mixed mean-reversion in high-beta Asia/tech pullbacks versus continuation in already-extended names.

CB-2026-10-10-1M official-20261010-1M monthly 2026-10-09 to 2026-11-09 FINANCIALS 35.0%SMALL_VALUE 35.0%SP500 30.0% Pending
Full portfolio
Financials Sector (XLF) 35.0% US Small-Cap Value (IWN) 35.0% S&P 500 (SPY) 30.0%
Rationale

SPY is near a 52-week high after a modest 5-session gain while 21-session breadth is weak (37.7% positive). Fed minutes show a September hike to 3.75%-4.00% and most participants expecting another increase by year-end amid 3.4% PCE and an ISM services prices index of 74.0. That mix favors the benchmark over stretched high-beta winners and over high-vol shock names unless a pullback is clearly transitory.

Audit and data Public Audit Packets For This Model