Investment Arena

Arena Methodology

Projected strategy definitions and execution disclosures for the AlphaSync Investment Arena.

How to Read the Experiments

These are comparison questions, not selected winners. Compare accounts within the same surface and season; check decision timing, completed bars, costs, and execution status before interpreting a return.

01 / Multi-asset Arena

LLM portfolios versus allocation rules

Equal weight, inverse volatility, and cross-asset trend provide transparent reference policies. Historical seasons contain deterministic replays only; a live LLM account has no pre-entry history to compare.

Open multi-asset Arena

02 / Single-asset Arena

LLM versus rules on one market

Compare the same instrument and season. An accepted target, a pending order, and a fill are distinct states; do not treat an unfilled target as executed exposure.

Open single-asset Arena

03 / Strategy Library

Backtest evidence versus Arena execution

The Library uses Strategyr for independent daily and 4-hour research. Arena uses Tradesimr for daily season execution. Four-hour crypto evidence is not ranked against the daily Arena.

Open Strategy Library

Execution and Season Disclosures

Completed-bar inputs: Not available in this registry release.
Next-eligible execution: Not available in this registry release.
Costs: Not available in this registry release.
Valuation: Not available in this registry release.
Season rules: Not available in this registry release.
LLM timing: Not available in this registry release.

Execution Assumptions

Initial capital: 1000000 USD
Decision data: Completed daily OHLC bars only.
Fill timing: Next eligible open after a completed-bar decision.
Deterministic rebalancing: Rebalance on a changed strategy target or when mark-to-market weight drifts by more than 5.0%. Weight tolerance: 5.0%
Transaction cost: 7 bps
Gross exposure cap: 1
Return basis: Synthetic price-return simulation, not total return.

Limitations: not modelled

These omissions are model limitations, not guarantees about live execution or investment outcomes.

ETF distributions and dividends
short borrow costs
FX carry
crypto funding
cash interest
bid-ask spread
market impact and slippage

LLM Workflow

Completed daily Yahoo OHLC bars plus locally cached FRED observations only.

Cutoff rule: Use only supplied observations at or before the declared context cutoff; do not rely on later information.
Response contract: JSON object with weights, decision_summary, and risk_flags.
Validation rules: All required weights, finite values, gross exposure cap 1, decision summary limit 240, and at most 3 risk flags.
Allowed instruments: 69, 118, 101, 114, 121, 32, 97, 99, 99, 111, 117, 110, 116, 32, 109, 117, 115, 116, 32, 114, 101, 116, 117, 114, 110, 32, 101, 120, 97, 99, 116, 108, 121, 32, 105, 116, 115, 32, 108, 105, 115, 116, 101, 100, 32, 97, 108, 108, 111, 119, 101, 100, 32, 115, 121, 109, 98, 111, 108, 115, 46

Completed market dataRead completed daily market bars; provisional current-UTC-date bars are excluded.
Local macro contextAttach available local FRED observations with their observation dates.
Dated decision packetVox produces an immutable completed-market and local macro packet under a declared cutoff.
ChatGPT Project decisionThe configured ChatGPT Project returns only the documented JSON allocation format.
Response validation and risk checksVox imports only a packet-bound response with allowed symbols, finite weights, gross exposure at or below one, and bounded public rationale fields.
Decision outcomePersist an accepted allocation or a no-decision record; invalid or unavailable responses retain the prior position.
Next eligible-bar executionTradesimr converts accepted weights to orders after the decision boundary and fills only on a later eligible completed bar.
Portfolio valuationTradesimr marks positions from completed bars and publishes public-safe ledger and quality records.

Projected LLM accounts

Account	Model	Provider	Allowed instruments	Decision cadence
ChatGPT Macro Balanced	chatgpt-project--macro-balanced	chatgpt_project_ui	SPY, TLT, GLD, EURUSD=X, BTC-USD, EEM, HYG, USO	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Risk Aware	chatgpt-project--risk-aware	chatgpt_project_ui	SPY, TLT, GLD, EURUSD=X, BTC-USD, EEM, HYG, USO	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Macro Balanced SPY	chatgpt-project--macro-balanced	chatgpt_project_ui	SPY	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Macro Balanced TLT	chatgpt-project--macro-balanced	chatgpt_project_ui	TLT	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Macro Balanced GLD	chatgpt-project--macro-balanced	chatgpt_project_ui	GLD	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Macro Balanced EUR/USD	chatgpt-project--macro-balanced	chatgpt_project_ui	EURUSD=X	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Macro Balanced BTC/USD	chatgpt-project--macro-balanced	chatgpt_project_ui	BTC-USD	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Macro Balanced EEM	chatgpt-project--macro-balanced	chatgpt_project_ui	EEM	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Macro Balanced HYG	chatgpt-project--macro-balanced	chatgpt_project_ui	HYG	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Macro Balanced USO	chatgpt-project--macro-balanced	chatgpt_project_ui	USO	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Risk Aware SPY	chatgpt-project--risk-aware	chatgpt_project_ui	SPY	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Risk Aware TLT	chatgpt-project--risk-aware	chatgpt_project_ui	TLT	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Risk Aware GLD	chatgpt-project--risk-aware	chatgpt_project_ui	GLD	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Risk Aware EUR/USD	chatgpt-project--risk-aware	chatgpt_project_ui	EURUSD=X	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Risk Aware BTC/USD	chatgpt-project--risk-aware	chatgpt_project_ui	BTC-USD	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Risk Aware EEM	chatgpt-project--risk-aware	chatgpt_project_ui	EEM	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Risk Aware HYG	chatgpt-project--risk-aware	chatgpt_project_ui	HYG	Only when Vox explicitly runs an LLM-enabled daily publication.
ChatGPT Risk Aware USO	chatgpt-project--risk-aware	chatgpt_project_ui	USO	Only when Vox explicitly runs an LLM-enabled daily publication.

Projected Strategy Definitions

buy_hold

Summary: Maintains a configured target exposure through the full input history.
Signal rule: Maintain the configured target exposure and rebalance only when the position differs from that target by more than the configured tolerance.
Position semantics: Positive values are long exposure, negative values are short exposure, and zero is flat.
Data requirements: [daily OHLC datetime]
Rebalance rule: The constant target is eligible after a completed source bar; an order is generated only when the configured tolerance is breached.

Parameters

Name	Value	Unit	Description
value	1	target exposure	Constant target exposure.

Technical provenancestrat_buy_and_hold_tgt_pos

ema_cross_adx

Summary: Uses EMA direction only when ADX indicates sufficient trend strength.
Signal rule: Go long when the fast EMA is above the slow EMA and ADX is at or above the threshold, go short for the reverse, and stay flat otherwise.
Position semantics: Positive values are long exposure, negative values are short exposure, and zero is flat.
Data requirements: [daily OHLC high daily OHLC low daily OHLC close]
Rebalance rule: A changed target is eligible after a completed source bar.

Parameters

Name	Value	Unit	Description
fast	20	bars	Fast EMA window.
slow	50	bars	Slow EMA window.
adx_n	14	bars	ADX window.
adx_threshold	20	ADX level	Minimum ADX level required for an active target.
target_size	1	target exposure	Absolute target exposure when a signal is active.

Technical provenancestrat_ema_cross_adx_tgt_pos

ema_cross_slope_confirm

Summary: Uses EMA alignment only when both EMA slopes confirm the same direction.
Signal rule: Go long when the fast EMA is above the slow EMA and both slopes are positive, go short when the reverse is true, and stay flat otherwise.
Position semantics: Positive values are long exposure, negative values are short exposure, and zero is flat.
Data requirements: [daily OHLC close]
Rebalance rule: A changed target is eligible after a completed source bar.

Parameters

Name	Value	Unit	Description
fast	20	bars	Fast EMA window.
slow	50	bars	Slow EMA window.
slope_lag	1	bars	Lag used to estimate EMA slope.
target_size	1	target exposure	Absolute target exposure when a signal is active.

Technical provenancestrat_ema_cross_slope_confirm_tgt_pos

rsi_revert

Summary: Uses classic RSI levels to open mean-reversion targets and close them near neutral.
Signal rule: Open long when RSI is oversold, open short when RSI is overbought, and close when RSI reaches the exit level.
Position semantics: Positive values are long exposure, negative values are short exposure, and zero is flat.
Data requirements: [daily OHLC close]
Rebalance rule: A changed target is eligible after a completed source bar.

Parameters

Name	Value	Unit	Description
n	14	bars	RSI window.
oversold	30	RSI level	Oversold threshold for long entry.
overbought	70	RSI level	Overbought threshold for short entry.
exit_level	50	RSI level	Neutral level used to close open targets.
target_size	1	target exposure	Absolute target exposure when a signal is active.

Technical provenancestrat_rsi_revert_tgt_pos

vol_target

Summary: Sets trend direction from price versus EMA and scales exposure by realized volatility.
Signal rule: Use price above the trend EMA for long direction, price below it for short direction, and cap size by target volatility versus realized volatility.
Position semantics: Positive values are long exposure, negative values are short exposure, and zero is flat.
Data requirements: [daily OHLC close]
Rebalance rule: A changed target is eligible after a completed source bar.

Parameters

Name	Value	Unit	Description
trend_n	20	bars	EMA window used for the directional trend filter.
rv_n	20	bars	Realized-volatility window.
vol_target	0.2	annualized volatility	Annualized target volatility.
max_leverage	1	target exposure	Maximum absolute target exposure.
annualization	252	bars per year	Annualization factor for realized volatility.

Technical provenancestrat_vol_target_tgt_pos

donchian_turtle

Summary: Uses a long Donchian breakout for entry and a shorter Donchian channel for exit.
Signal rule: Enter long above the prior entry channel high, enter short below the prior entry channel low, and exit on the opposite shorter exit channel.
Position semantics: Positive values are long exposure, negative values are short exposure, and zero is flat.
Data requirements: [daily OHLC high daily OHLC low daily OHLC close]
Rebalance rule: A changed target is eligible after a completed source bar.

Parameters

Name	Value	Unit	Description
entry_n	55	bars	Donchian entry window.
exit_n	20	bars	Donchian exit window.
target_size	1	target exposure	Absolute target exposure when a signal is active.

Technical provenancestrat_donchian_turtle_tgt_pos

bollinger_revert

Summary: Uses Bollinger band touches to open mean-reversion targets and closes near the middle band.
Signal rule: Open long at or below the lower band, open short at or above the upper band, and close when price returns to the middle band.
Position semantics: Positive values are long exposure, negative values are short exposure, and zero is flat.
Data requirements: [daily OHLC close]
Rebalance rule: A changed target is eligible after a completed source bar.

Parameters

Name	Value	Unit	Description
n	20	bars	Bollinger window.
k	2	standard deviations	Bollinger width multiplier.
target_size	1	target exposure	Absolute target exposure when a signal is active.

Technical provenancestrat_bollinger_revert_tgt_pos

regime_switch

Summary: Switches among trend, mean-reversion, and flat exposure using ADX and realized volatility state.
Signal rule: Use EMA trend targets in strong non-high-volatility trend regimes, Bollinger reversion targets in weak non-high-volatility regimes, and flat exposure otherwise.
Position semantics: Positive values are long exposure, negative values are short exposure, and zero is flat.
Data requirements: [daily OHLC high daily OHLC low daily OHLC close]
Rebalance rule: A changed target is eligible after a completed source bar.

Parameters

Name	Value	Unit	Description
fast	20	bars	Fast EMA window for the trend regime.
slow	50	bars	Slow EMA window for the trend regime.
adx_n	14	bars	ADX window.
rv_n	20	bars	Realized-volatility window.
bb_n	20	bars	Bollinger window for the reversion regime.
bb_k	2	standard deviations	Bollinger width multiplier for the reversion regime.
trend_adx_threshold	25	ADX level	ADX level at or above which the trend regime is allowed.
revert_adx_threshold	18	ADX level	ADX level at or below which the reversion regime is allowed.
high_vol_threshold	0.4	annualized volatility	Maximum realized volatility allowed for active regimes.
breadth_long_threshold		breadth value	Disabled by default; no lower breadth bound is applied.
breadth_short_threshold		breadth value	Disabled by default; no upper breadth bound is applied.
annualization	252	bars per year	Annualization factor for realized volatility.
target_size	1	target exposure	Absolute target exposure when a signal is active.

Technical provenancestrat_regime_switch_tgt_pos

equal_weight_rebalance

Summary: Allocates equally across eligible daily instruments on a fixed cadence.
Signal rule: At each rebalance signal, divide the allowed gross exposure equally across eligible assets and hold the resulting targets until the next complete shared-asset open by default.
Position semantics: Weights are long-only portfolio fractions; unallocated equity remains cash.
Data requirements: [daily OHLC open daily OHLC high daily OHLC low daily OHLC close asset identifier date]
Rebalance rule: Signals use completed bars and targets execute at the next complete shared-asset open by default.

Parameters

Name	Value	Unit	Description
date_col	date	column name	Date column name.
asset_col	asset	column name	Asset identifier column name.
open_col	open	column name	Daily open-price column name.
high_col	high	column name	Daily high-price column name.
low_col	low	column name	Daily low-price column name.
close_col	close	column name	Daily close-price column name.
rebalance_n	21	bars	Completed daily bars between rebalance signals.
min_obs	1	observations	Minimum observed closes required for eligibility.
gross_exposure	1	portfolio weight	Maximum long gross exposure; residual remains cash.
weight_cap	1	portfolio weight	Maximum target weight for one asset.
rebalance_calendar	shared	calendar mode	Use complete shared asset boundaries; use available only for an explicit variable-universe schedule.

Technical provenancestrat_equal_weight_rebalance_target_weights

inverse_volatility_allocation

Summary: Allocates across eligible daily instruments in inverse proportion to realized volatility.
Signal rule: At each rebalance signal, allocate the allowed gross exposure by inverse realized volatility and hold targets until the next complete shared-asset open by default.
Position semantics: Weights are long-only portfolio fractions; unallocated equity remains cash.
Data requirements: [daily OHLC open daily OHLC high daily OHLC low daily OHLC close asset identifier date]
Rebalance rule: Signals use completed bars and targets execute at the next complete shared-asset open by default.

Parameters

Name	Value	Unit	Description
date_col	date	column name	Date column name.
asset_col	asset	column name	Asset identifier column name.
open_col	open	column name	Daily open-price column name.
high_col	high	column name	Daily high-price column name.
low_col	low	column name	Daily low-price column name.
close_col	close	column name	Daily close-price column name.
vol_n	20	bars	Realized-volatility lookback in daily returns.
min_obs	20	observations	Minimum observed closes required for eligibility.
annualization	252	bars per year	Annualization factor for realized volatility.
rebalance_n	21	bars	Completed daily bars between rebalance signals.
gross_exposure	1	portfolio weight	Maximum long gross exposure; residual remains cash.
weight_cap	0.4	portfolio weight	Maximum target weight for one asset.
rebalance_calendar	shared	calendar mode	Use complete shared asset boundaries; use available only for an explicit variable-universe schedule.

Technical provenancestrat_inverse_volatility_allocation_target_weights

cross_asset_trend_allocation

Summary: Allocates across assets with positive medium-term momentum and otherwise remains in cash.
Signal rule: At each rebalance signal, include only assets with positive completed-bar momentum and allocate by equal or inverse-volatility weights for the next complete shared-asset open by default.
Position semantics: Weights are long-only portfolio fractions; unallocated equity remains cash.
Data requirements: [daily OHLC open daily OHLC high daily OHLC low daily OHLC close asset identifier date]
Rebalance rule: Signals use completed bars and targets execute at the next complete shared-asset open by default.

Parameters

Name	Value	Unit	Description
date_col	date	column name	Date column name.
asset_col	asset	column name	Asset identifier column name.
open_col	open	column name	Daily open-price column name.
high_col	high	column name	Daily high-price column name.
low_col	low	column name	Daily low-price column name.
close_col	close	column name	Daily close-price column name.
trend_n	126	bars	Medium-term momentum lookback.
vol_n	20	bars	Realized-volatility lookback in daily returns.
min_obs	126	observations	Minimum observed closes required for eligibility.
annualization	252	bars per year	Annualization factor for realized volatility.
volatility_scale	true	logical	Use inverse realized-volatility scaling after the trend filter.
rebalance_n	21	bars	Completed daily bars between rebalance signals.
gross_exposure	1	portfolio weight	Maximum long gross exposure; residual remains cash.
weight_cap	0.4	portfolio weight	Maximum target weight for one asset.
rebalance_calendar	shared	calendar mode	Use complete shared asset boundaries; use available only for an explicit variable-universe schedule.

Technical provenancestrat_cross_asset_trend_allocation_target_weights

Universe and Instruments

SPY

SPDR S&P 500 ETF Trust

Asset ID: 1
Asset class: equity
Instrument type: etf
Quote currency: USD
Market calendar: XNYS
Arena role: US equity benchmark
Price data rule: Completed daily OHLC bars only.
Source: Investdatar completed daily OHLC

SPDR S&P 500 ETF Trust, used as the Arena's US equity benchmark.

TLT

iShares 20+ Year Treasury Bond ETF

Asset ID: 2
Asset class: fixed_income
Instrument type: etf
Quote currency: USD
Market calendar: XNYS
Arena role: Duration and rates exposure
Price data rule: Completed daily OHLC bars only.
Source: Investdatar completed daily OHLC

iShares 20+ Year Treasury Bond ETF, used as the Arena's Duration and rates exposure.

GLD

SPDR Gold Shares

Asset ID: 3
Asset class: commodity
Instrument type: etf
Quote currency: USD
Market calendar: XNYS
Arena role: Gold commodity exposure
Price data rule: Completed daily OHLC bars only.
Source: Investdatar completed daily OHLC

SPDR Gold Shares, used as the Arena's Gold commodity exposure.

EURUSD=X

Euro / U.S. Dollar

Asset ID: 4
Asset class: foreign_exchange
Instrument type: spot_fx
Quote currency: USD
Market calendar: FX_24_5
Arena role: EUR versus USD exposure
Price data rule: Completed daily OHLC bars only.
Source: Investdatar completed daily OHLC

Euro / U.S. Dollar, used as the Arena's EUR versus USD exposure.

BTC-USD

Bitcoin / U.S. Dollar

Asset ID: 5
Asset class: cryptocurrency
Instrument type: spot_crypto
Quote currency: USD
Market calendar: CRYPTO_24_7
Arena role: Digital-asset exposure
Price data rule: Completed daily OHLC bars only.
Source: Investdatar completed daily OHLC

Bitcoin / U.S. Dollar, used as the Arena's Digital-asset exposure.

EEM

iShares MSCI Emerging Markets ETF

Asset ID: 6
Asset class: equity
Instrument type: etf
Quote currency: USD
Market calendar: XNYS
Arena role: Emerging-market equity exposure
Price data rule: Completed daily OHLC bars only.
Source: Investdatar completed daily OHLC

iShares MSCI Emerging Markets ETF, used as the Arena's Emerging-market equity exposure.

HYG

iShares iBoxx $ High Yield Corporate Bond ETF

Asset ID: 7
Asset class: fixed_income
Instrument type: etf
Quote currency: USD
Market calendar: XNYS
Arena role: Credit-cycle and risk-premium exposure
Price data rule: Completed daily OHLC bars only.
Source: Investdatar completed daily OHLC

iShares iBoxx $ High Yield Corporate Bond ETF, used as the Arena's Credit-cycle and risk-premium exposure.

USO

United States Oil Fund

Asset ID: 8
Asset class: commodity
Instrument type: etf
Quote currency: USD
Market calendar: XNYS
Arena role: Energy and inflation-sensitive commodity exposure
Price data rule: Completed daily OHLC bars only.
Source: Investdatar completed daily OHLC

United States Oil Fund, used as the Arena's Energy and inflation-sensitive commodity exposure.