AMM divergence versus holding

How does a fee-free 50/50 constant-product liquidity position differ from holding its original two assets after a token-price change?

This experiment changes declared financial assumptions and calculates their consequences. The starting figures are illustrative; the model does not inspect your accounts or require a live market feed. Run the question free, compare an alternative and inspect the calculation receipt.

What the published starting values calculate.

Pool value minus holding value: -857.86 USD.

The fee-free constant-product position is compared with holding the same initial assets. Relative divergence is not necessarily an absolute loss: both positions may gain in USD. The path is an entered scenario, not a forecast.

Cash-funding feasibility is not assessed. Modelled value, interest or a conditional fee quote does not establish available settled cash.

The illustrative date origin is 1 January 2026; the recorded seed is 42 and the model is yieldguard-experiments-1.2.0. Dates and values are scenario inputs, not current observations. Readable percentages may be rounded; the evidence file preserves the kernel values.

Starting assumptions and units.

What the result establishes and leaves open.

The calculation is conditional on declared input values, model conventions and horizon. Reproducing the result checks the disclosed calculation path; it does not establish forecasting accuracy or the suitability of a real treasury decision.

• Illustrative user assumptions; no live market data or calibrated forecast.
• Results apply only to the declared mechanisms, horizon and search bounds.
• Exact declared model: x*y=k, initially equal USD values, one fixed-USD quote asset, immediate arbitrage, constant liquidity ownership and zero trading fees or incentives. This is not concentrated liquidity or a protocol return forecast.
• Initial investment must be positive and the final/initial price ratio must be between 0.01 and 100. The displayed price path is an assumed linear interpolation. Values are not settled cash. Source: https://app.uniswap.org/whitepaper.pdf
• No market refresh, AI calculation, trade execution or forecast calibration.
• Receipts are content-addressed local calculation records, not externally signed attestations.
• Deterministic calculation: the seed is recorded for the common receipt contract; this experiment samples no random disturbances.
• Illustrative assumptions; no live data or forecast calibration.
• Timeline is a value series, not settlement availability. Cash-funding feasibility is not assessed.

The next useful question.

Does flat assumed yield cover entry, gas and exit costs over the chosen holding period?

Keep a recurring question under review.

In the Laboratory, edit the assumptions and compare the result. Planned Macrovision would connect approved organisational evidence, material changes, assigned exception review and preserved decisions. AI prepares and explains; approved models calculate; your team decides.

Planned — not accepting paid engagements. The public Macrovision walkthrough follows a synthetic late-receipt case. The walkthrough demonstrates intended work and does not activate customer monitoring or imply that every Lab question is a supported paid workflow.

Financial assumptions are not automatically sent into an enquiry, analytics or a shared result URL. Any later transfer to a hosted workspace must be explicit.