This experiment tries to replay a public Polymarket account's filled trade
ledger inside the normal backtesting framework. It was built around the
@beffer45 profile and wallet
0xe29aff6a6ae1e1d6a3a1c4c904f2957afa98cda0, using a hard-coded snapshot of
public trade rows from the Polymarket data API.
The goal is not to present a profitable strategy. The goal is to ask a stricter realism question:
If we know the trades a public account took, can the backtester reproduce the
same portfolio by replaying those decisions against the historical book?
The answer from this experiment is no, not exactly. That failure is still useful because it shows where a filled account ledger stops being a strategy and starts being a record of events that already happened inside the market data.
Runner And Notebook
The Python runner is:
uv run python backtests/polymarket_beffer45_trade_replay_telonex.py
The notebook wrapper is:
backtests/polymarket_beffer45_trade_replay_telonex.ipynb
The notebook calls the same runner and stores the latest generated HTML summary inside the notebook output. A user who downloads the repository can open the notebook and inspect the last saved output without rerunning the backtest, subject to the normal notebook trust and renderer behavior of their local Jupyter or GitHub viewer.
The runner writes:
output/polymarket_beffer45_trade_replay_telonex_summary.htmloutput/polymarket_beffer45_trade_replay_telonex_comparison.csv
Those output files are generated artifacts. The committed notebook is the portable way to keep the last HTML view attached to the example.
What The Strategy Does
strategies.account_trade_replay.BookAccountTradeReplayStrategy receives the
hard-coded public ledger rows and schedules one order at each public trade
timestamp.
Important details:
- The replay uses
BookReplayover Telonex L2 order book data. - It submits limit IOC orders at the public ledger price.
- It does not manufacture fills.
- It does not bypass the Nautilus execution model.
- It uses the same report and portfolio machinery as the other public runners.
- Sells are treated as reduce-only attempts so the strategy cannot invent short exposure that the public ledger did not show.
This is intentionally strict. A forced-fill ledger simulator would reproduce the CSV by construction, but it would not test the backtesting engine's execution model. The point of this runner is to let the historical book accept or reject each attempted replay order.
Why Exact Reproduction Fails
A public filled-trade feed is not an order-management history. It omits several facts that matter for realistic replay.
First, the original user's trades are already reflected in the historical order book and trade stream. If the account crossed the spread, consumed depth, or caused a book update, that effect is part of the data the backtest later replays. Re-submitting the same account's order into that same book is not a clean counterfactual. It can double-count the interaction with liquidity, or it can miss the liquidity that the original order already consumed.
Second, the public row timestamp is a fill observation, not necessarily the decision time. A copy trader only sees the trade after it is public. Even a zero-latency replay at the fill timestamp is already an optimistic assumption for copying. A real copier would submit later and face a different queue, spread, and available depth.
Third, maker fills cannot be reconstructed from filled trade rows alone. A maker fill depends on when the resting order was placed, how it was amended or canceled, and where it sat in queue. The public account trade row may show that the account bought or sold, but it does not provide the missing order lifecycle.
Fourth, Polymarket markets can be thin and discontinuous. Small differences in timestamp handling, minimum size checks, trade-tick availability, queue position, or book reconstruction can decide whether an IOC order fills at all. That is not noise to hide. It is exactly the kind of realism gap this framework should surface.
Finally, settlement and mark-to-market accounting are separate from order replay. The public profile can report settled account PnL over its own portfolio history. The backtest report only settles instruments when the replay horizon and metadata make that outcome available to the report without leaking future information into the strategy.
Copy-Trading Interpretation
This experiment is a useful warning against a naive copy-trading claim:
"If a public account made these trades, replaying or copying the trades should
produce the same portfolio."
That statement is usually false.
A public fill is an after-the-fact event. It is not a signal that existed before the market moved, and it is not a complete description of the order that generated the fill. Copying a public wallet from a trade feed therefore has at least three disadvantages:
- the copied order is later than the original order
- the original order's market impact is already embedded in the observed book
- the copier does not know the original queue position, passive order history, cancellations, or hidden intent
This does not prove that every account-following strategy is impossible. A strategy that models reporting delay, liquidity decay, market selection, and post-fill price drift could still be worth studying. What this runner shows is that a hard-coded public ledger is not itself a tradable strategy, and exact portfolio reproduction should not be expected from L2 replay.
External Source Check
The realism caveats above are grounded in the public mechanics documented by Polymarket and NautilusTrader:
- Polymarket documents separate API surfaces: the public Data API covers user positions, trades, activity, holders, open interest, leaderboards, and builder analytics, while the CLOB API covers orderbook data, pricing, and trading operations. Trading endpoints require authentication. Source: Polymarket API Introduction.
- Polymarket describes the CLOB as an offchain matching system with onchain settlement, where orders are signed and matched trades settle on Polygon. Source: Polymarket Trading Overview.
- Polymarket's order docs say orders are expressed as limit orders. GTC and GTD orders rest on the book, while FOK and FAK execute against resting liquidity immediately. The same page documents rejections for minimum size, insufficient balance, tick-size mismatches, and unfilled FOK orders. Source: Polymarket Create Order.
- Polymarket's authenticated user WebSocket examples include order lifecycle
fields such as
original_size,size_matched,created_at,status, andorder_type, plus trade fields such astaker_order_id,maker_orders,trader_side, andtimestamp. A public filled-trade snapshot is therefore not equivalent to the complete private order lifecycle. Source: Polymarket User Channel. - Polymarket's trade history docs distinguish open order fields from trade fields and show that trades have match/finality statuses and maker-side detail. This supports treating the public ledger as a fill record, not as a full order-management log. Source: Polymarket Cancel Order / Trade History.
- NautilusTrader documents that L2/L3 book data drives book state,
TradeTickcan trigger matching, and historical order book/trade data is immutable in backtesting. It also documents that order book deltas are required forL2_MBP, and that trade ticks with L2 book data provide execution evidence rather than automatically filling every order. Source: NautilusTrader Backtesting.
The copy-trading interpretation is an inference from those documented mechanics: if the historical replay data already contains the original account's fills and book updates, then replaying those same fills as new orders against that history is not the same counterfactual as having submitted the original orders with the original queue position and timing.
Observed Result
The checked-in snapshot contains 153 public trade rows across 86 instruments. On the Telonex replay used for this experiment, the strict replay filled 72 of those 153 attempted orders.
The latest run reported:
- loaded instruments:
86 / 86 - engine fills:
72 / 153 - ledger cash PnL on loaded instruments:
-223.8793596488217363403119430 USDC - instruments with resolved-outcome metadata:
86 - instruments with settlement applied by report:
44 - ledger settlement PnL using resolved metadata:
50.44854735117825935968805694 USDC - backtest report PnL on loaded instruments:
-16.2740783200000076046 USDC - delta between report PnL and ledger settlement metadata:
-66.72262567117826696428805694 USDC
These numbers should be read as diagnostic output, not as an account audit. The important result is the divergence: many public ledger fills cannot be replayed as fresh IOC orders against the already-realized historical book.
Latest Terminal Output
──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Market Book Events Fills Qty AvgPx Notional PnL (USDC) Return MaxDD Sharpe Sortino PF Coverage
──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
bitcoin-up-or-down-april-26-2026-4am-et:Up 101833 1 14.01 0.3400 4.76 -4.6995 -0.47% -1.18% n/a -15.87 0.88 +100.00%
btc-updown-15m-1777222800:Up 61180 2 10.45 0.2400 2.51 +0.6700 +0.07% -0.02% n/a n/a 2.17 +100.00%
btc-updown-15m-1777222800:Down 41584 1 92.87 0.0100 0.93 -0.9103 -0.09% -0.09% n/a -15.87 0.74 +100.00%
btc-updown-15m-1777249800:Down 71869 2 101.00 0.0100 1.01 -0.9900 -0.10% -0.10% n/a -15.87 0.78 +44.08%
btc-updown-5m-1776990300:Up 69850 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1776990600:Up 108375 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1776990900:Up 51924 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777056600:Up 171695 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777057500:Up 113222 1 52.82 0.6300 33.28 -33.0298 -3.30% -3.38% n/a -15.87 0.81 +100.00%
btc-updown-5m-1777119000:Up 127051 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777119000:Down 127460 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777119300:Up 129809 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777119900:Up 93564 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777119900:Down 93557 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777126800:Up 90322 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777129200:Down 121662 1 31.29 0.0100 0.31 -0.3067 -0.03% -2.98% n/a n/a 1.01 +100.00%
btc-updown-5m-1777129500:Up 99228 1 5.00 0.1400 0.70 -0.6880 -0.07% -0.15% n/a -15.87 0.90 +100.00%
btc-updown-5m-1777134000:Up 124347 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777134300:Down 144982 1 21.37 0.8100 17.31 +4.1261 +0.41% -0.35% n/a n/a 1.35 +100.00%
btc-updown-5m-1777135500:Up 170536 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777135800:Up 162715 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777137000:Down 132476 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777159800:Down 148439 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777170300:Up 112558 1 10.38 0.0800 0.83 -0.8154 -0.08% -0.58% n/a -15.87 0.98 +100.00%
btc-updown-5m-1777170600:Up 115574 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777171200:Up 94273 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777171500:Down 118028 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777171800:Up 113832 1 20.89 0.4019 8.40 +12.5968 +1.26% -0.84% n/a n/a 1.29 +100.00%
btc-updown-5m-1777172100:Up 115178 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777172100:Down 115159 1 35.71 0.0100 0.36 -0.3500 -0.04% -0.14% n/a -15.87 0.76 +100.00%
btc-updown-5m-1777172400:Up 93492 1 8.51 0.1100 0.94 -0.9193 -0.09% -0.12% n/a -15.87 0.63 +100.00%
btc-updown-5m-1777173000:Up 102545 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777173600:Up 117590 1 8.95 0.5400 4.83 -4.7899 -0.48% -0.69% n/a -15.87 0.90 +100.00%
btc-updown-5m-1777173900:Up 104785 1 11.67 0.0800 0.93 -0.9166 -0.09% -0.12% n/a -15.87 0.85 +100.00%
btc-updown-5m-1777173900:Down 104791 1 10.91 0.4376 4.77 +6.1874 +0.62% -0.09% n/a n/a 1.27 +100.00%
btc-updown-5m-1777174200:Up 119613 1 15.45 0.0600 0.93 -0.9095 -0.09% -0.09% n/a -15.87 0.56 +100.00%
btc-updown-5m-1777174500:Up 135045 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777174500:Down 135071 1 16.45 0.5500 9.05 +7.4834 +0.75% -0.55% n/a n/a 1.08 +100.00%
btc-updown-5m-1777174800:Down 100761 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777175100:Up 116283 1 15.38 0.2700 4.15 -4.0913 -0.41% -1.51% n/a -15.87 0.94 +100.00%
btc-updown-5m-1777175100:Down 116383 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777175400:Up 122794 1 29.36 0.1300 3.82 -3.7504 -0.38% -2.16% n/a -15.87 0.97 +100.00%
btc-updown-5m-1777175700:Up 134559 3 30.05 0.4537 13.63 +16.5576 +1.66% -1.92% n/a n/a 1.09 +100.00%
btc-updown-5m-1777175700:Down 134564 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777176000:Down 104622 1 8.51 0.0900 0.77 +7.7566 +0.78% -0.62% n/a n/a 1.20 +100.00%
btc-updown-5m-1777176600:Up 116945 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777177800:Up 92877 1 33.50 0.1400 4.69 -4.6097 -0.46% -0.64% n/a -15.87 0.83 +100.00%
btc-updown-5m-1777178100:Down 91779 1 16.66 0.0400 0.67 -0.6536 -0.07% -0.24% n/a -15.87 0.91 +100.00%
btc-updown-5m-1777178400:Down 96229 2 14.42 0.1350 1.95 +0.2499 +0.02% -0.04% n/a n/a 1.13 +100.00%
btc-updown-5m-1777178700:Down 91725 2 58.14 0.0240 1.40 -1.3691 -0.14% -0.19% n/a -15.87 0.66 +100.00%
btc-updown-5m-1777179600:Up 95365 3 67.01 0.1400 9.38 +0.8309 +0.08% -0.18% n/a n/a 1.26 +100.00%
btc-updown-5m-1777179900:Down 90094 2 62.01 0.0200 1.24 -0.5960 -0.06% -0.06% n/a -15.87 0.23 +100.00%
btc-updown-5m-1777180200:Up 83488 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777180500:Up 121917 2 92.94 0.3050 28.35 +26.7309 +2.67% -1.23% n/a n/a 1.33 +100.00%
btc-updown-5m-1777180800:Up 100572 3 90.27 0.1588 14.34 -14.1513 -1.42% -1.50% n/a -15.87 0.67 +100.00%
btc-updown-5m-1777181100:Up 93236 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777181100:Down 93238 2 31.99 0.3325 10.64 -10.5400 -1.05% -1.06% n/a -15.87 0.79 +100.00%
btc-updown-5m-1777181400:Up 115139 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777181700:Up 99565 1 20.46 0.4700 9.62 -9.5164 -0.95% -0.96% n/a -15.87 0.85 +100.00%
btc-updown-5m-1777183500:Up 92191 2 12.51 0.0900 1.13 -0.4808 -0.05% -0.05% n/a -15.87 0.48 +100.00%
btc-updown-5m-1777183500:Down 92202 2 27.21 0.7750 21.09 +1.5917 +0.16% -0.14% n/a n/a 1.37 +100.00%
btc-updown-5m-1777189500:Up 104196 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777189800:Down 88289 1 58.36 0.0800 4.67 -4.5829 -0.46% -0.52% n/a -15.87 0.61 +100.00%
btc-updown-5m-1777222500:Up 89040 1 46.47 0.0200 0.93 -0.9112 -0.09% -0.12% n/a -15.87 0.61 +100.00%
btc-updown-5m-1777223700:Up 108963 2 20.33 0.5300 10.77 +1.7277 +0.17% -0.29% n/a n/a 1.19 +100.00%
btc-updown-5m-1777224000:Up 114377 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777239600:Up 92689 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777239600:Down 92689 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777239900:Up 114596 2 44.85 0.1520 6.82 -2.0408 -0.20% -0.24% n/a -15.87 0.41 +100.00%
btc-updown-5m-1777239900:Down 114859 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777240200:Up 107984 2 17.90 0.5250 9.40 +0.0010 +0.00% -0.03% n/a n/a 1.00 +100.00%
btc-updown-5m-1777240200:Down 107981 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
btc-updown-5m-1777240500:Down 127097 1 116.02 0.1659 19.25 -18.9275 -1.89% -7.09% n/a -15.87 0.99 +100.00%
btc-updown-5m-1777240800:Up 135036 2 102.22 0.0100 1.02 -1.0020 -0.10% -0.22% n/a -15.87 0.55 +100.00%
btc-updown-5m-1777240800:Down 135000 1 23.02 0.8600 19.80 +3.2784 +0.33% -1.00% n/a n/a 1.07 +100.00%
btc-updown-5m-1777241100:Up 67900 2 47.76 0.8400 40.12 +2.9905 +0.30% -0.23% n/a n/a 1.24 +100.00%
btc-updown-5m-1777241100:Down 67771 1 77.69 0.0400 3.11 -3.0481 -0.30% -0.47% n/a -15.87 0.76 +100.00%
btc-updown-5m-1777241400:Down 102409 2 15.72 0.0992 1.56 -1.5308 -0.15% -0.55% n/a -15.87 0.93 +100.00%
btc-updown-5m-1777249800:Down 151613 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +55.59%
cricipl-del-pun-2026-04-25:Punjab Kings 182587 2 14.98 0.5645 8.46 +6.5315 +0.65% -0.29% n/a n/a 1.23 +100.00%
cricipl-luc-kol-2026-04-26:Lucknow Super Giants 105152 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +57.62%
cricipl-luc-kol-2026-04-26:Kolkata Knight Riders 151976 2 80.24 0.8950 71.82 +3.7953 +0.38% -0.60% n/a n/a 1.22 +59.76%
cricipl-raj-sun-2026-04-25:Rajasthan Royals 31606 1 8.11 0.1200 0.97 +0.0214 -0.09% -0.14% n/a -15.87 0.87 +100.00%
cricipl-roy-guj-2026-04-24:Royal Challengers Bangalore 261231 2 56.51 0.7945 44.89 +11.7258 +1.17% -0.65% n/a n/a 1.17 +100.00%
eth-updown-5m-1777170900:Up 39922 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
eth-updown-5m-1777171200:Down 39194 0 0.00 n/a 0.00 +0.0000 +0.00% -0.00% n/a n/a n/a +100.00%
──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
TOTAL 9385899 72 1808.34 0.2556 462.27 -16.2741 -1.73% -9.28% -0.20 -0.27 1.00 +97.87%
──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Portfolio run stats: Iterations: 9,500,965 | Events: 410 | Orders: 153 | Positions: 48 | Elapsed: 260867.127s
Portfolio return stats: Sharpe Ratio (252 days): -10.87 | Sortino Ratio (252 days): -10.2 | Profit Factor: 0 | Risk Return Ratio: -0.6849 | Returns Volatility (252 days): 1.017 | Average (Return): -0.04387
Portfolio PnL stats (USDC.e): PnL (total): -148.5 | PnL% (total): -14.85 | Win Rate: 0.9167 | Expectancy: 1.164 | Avg Winner: 1.341 | Avg Loser: -0.7787
WARNING: bitcoin-up-or-down-april-26-2026-4am-et:Up: Replay selection is explicitly curated from named markets and may exclude cancelled, delisted, or zero-liquidity markets.
WARNING: bitcoin-up-or-down-april-26-2026-4am-et:Up: No portfolio-level drawdown or daily-loss circuit breaker is configured for this run.
WARNING: cricipl-del-pun-2026-04-25:Punjab Kings: Settlement outcome exists after the replay window; keeping mark-to-market PnL instead of resolved settlement because resolution was not observable by 2026-04-27T00:46:48.152000+00:00 (observable at 2026-05-02T00:00:00+00:00).
WARNING: cricipl-luc-kol-2026-04-26:Lucknow Super Giants: Settlement outcome exists after the replay window; keeping mark-to-market PnL instead of resolved settlement because resolution was not observable by 2026-04-27T00:46:48.152000+00:00 (observable at 2026-05-03T00:00:00+00:00).
WARNING: cricipl-luc-kol-2026-04-26:Kolkata Knight Riders: Settlement outcome exists after the replay window; keeping mark-to-market PnL instead of resolved settlement because resolution was not observable by 2026-04-27T00:46:48.152000+00:00 (observable at 2026-05-03T00:00:00+00:00).
WARNING: cricipl-raj-sun-2026-04-25:Rajasthan Royals: Settlement outcome exists after the replay window; keeping mark-to-market PnL instead of resolved settlement because resolution was not observable by 2026-04-27T00:46:48.152000+00:00 (observable at 2026-05-02T00:00:00+00:00).
WARNING: cricipl-roy-guj-2026-04-24:Royal Challengers Bangalore: Settlement outcome exists after the replay window; keeping mark-to-market PnL instead of resolved settlement because resolution was not observable by 2026-04-27T00:46:48.152000+00:00 (observable at 2026-05-01T00:00:00+00:00).
Summary report saved to /Users/evankolberg/prediction-market-backtesting/output/polymarket_beffer45_trade_replay_telonex_summary.html
Backtest vs hard-coded ledger comparison
loaded instruments: 86 / 86
engine fills: 72 / 153 ledger trades
ledger cash PnL on loaded instruments: -223.8793596488217363403119430 USDC
instruments with resolved-outcome metadata: 86
instruments with settlement applied by report: 44
ledger settlement PnL using resolved metadata: 50.44854735117825935968805694 USDC
backtest report PnL on loaded instruments: -16.2740783200000076046 USDC
delta report - ledger settlement metadata: -66.72262567117826696428805694 USDC
comparison CSV: output/polymarket_beffer45_trade_replay_telonex_comparison.csv
summary HTML: output/polymarket_beffer45_trade_replay_telonex_summary.html
Total wall time: 310.47s
(base) evankolberg@Evans-MacBook-Pro-2 prediction-market-backtesting %How To Use This Experiment
Use this runner as a regression and realism harness:
- to check that account-ledger replay orders go through the same execution path as other strategies
- to inspect which public fills are accepted or rejected by the historical book
- to compare account-level public ledger accounting against backtest accounting
- to test notebook output persistence for HTML summary reports
- to reason about copy-trading assumptions without changing core engine behavior
Do not use it as proof that the account's profile PnL is wrong, and do not use it as proof that the backtester should force fills to match a public portfolio. The mismatch is the point. It identifies information that a filled trade ledger does not contain and market impact that historical data has already absorbed.