Portfolio Backtests: Include Carry and Shared Risk

Portfolio backtests: Synchronize portfolio sleeves and charge declared financing costs before comparing combined returns.

A multisleeve portfolio discussion raised omitted financing and changing contribution questions. The reviewed Reddit post was published August 5, 2026. Its author results were not independently reproduced, and the available comment tree remains incomplete. The exercise below is our original research proposal, not a replication of the post.

Portfolio diversification must be tested on a shared calendar. Independently shuffling each sleeve creates combinations that never occurred together and can manufacture protection. Start with one row per timestamp containing every sleeve, cash and net equity. Preserve inactive periods and mark missing observations explicitly.

Portfolio backtests: exercise: a financing bridge

Use an invented portfolio with starting equity of $10,000 and $20,000 exposure. Suppose its price gains contribute $300 over thirty calendar days. An assumed $10,000 debit at 12% annual interest costs about $98.63 under simple ACT/365 arithmetic. Before fees, equity gains are about $201.37. Charging 12% on all exposure would answer a different stress question; label it separately.

Add signed funding payments in actual dollars and record whether a positive input means a debit or a credit. For partial positions, compute the supplied payment from actual exposure. Do not silently scale a payment twice. Weekends still occupy calendar time even when the price series contains no new session.

Inspect contributions together

Remove one sleeve at a time while declaring whether its capital stays in cash or is redistributed. Those experiments are different portfolios. During the original worst months, inspect each sleeve's contribution and concurrent exposure. Two defensive sleeves falling together need not explain the account's loss if other holdings dominate its dollars at risk.

A regression test should conserve starting cash plus realized gains, marks and financing into ending equity. A second test should preserve a deliberately synchronized loss when resampling rows. A third should reject financing when elapsed time is unavailable. These checks are stronger than a correlation label because they expose precisely which accounting assumption changes the answer.

Additional validation

Write the financing convention in the table header, including its day basis and sign. Keep gross market return, debit interest, other carry, funding and transaction costs as separate dollar columns. This lets another reviewer change one assumption without rebuilding unrelated performance calculations. Check both a funded cash account and a leveraged account as separate fixtures.

Evidence and workspace use

Primary reference: Alpaca Paper Trading provides background for this experiment, not validation of the Reddit strategy. Treat invented numbers above as arithmetic fixtures. Preserve the dataset, configuration and timestamp conventions with the output so another researcher can inspect the same calculation.

Find the offline workflow through Tradewink Advisory Research. Choose a synthetic sample before importing offline JSON. Start in an advisory research workspace: save a small input fixture, write the expected result by hand, then compare the implemented output. Expand the dataset only after the fixture balances. Keep unresolved assumptions visible in the report and retain adverse outcomes. Synthetic research output does not establish live execution behavior. For sizing background, read position sizing; for broader context, see risk management.

Continue the research

Bitcoin Trading Features: Audit Redundant Inputs and Backtest Reproducibility: Build an Audit Checklist develop adjacent checks. Use a small synthetic fixture first; these educational diagnostics do not validate a trading edge or change account settings.