A performance ranking is meaningful only when the assets are measured over comparable periods and in the same reporting unit. Fix the start and end before interpreting the result. Historical percentages describe the selected observations, not a promised return for the next period.
A changed baseline can change the ranking without changing the underlying price history.
BTKWIN.COM · EDUCATIONAL EXAMPLEState the question before choosing the dates
Decide whether you are comparing market prices, a hypothetical portfolio or actual investment results. Each requires different evidence. A chart return can use a fixed unit, while a personal result needs cash flows, quantities and costs.
Write the period and selection rule before inspecting winners. Choosing a start date at one asset’s low after seeing the outcome can exaggerate its relative performance. A ranking should preserve why the baseline was selected.
Keep the original version dated. If you later examine another period, label it a separate comparison rather than replacing the first one. This makes the influence of the chosen window visible.
Use a common reporting currency
An asset can have different returns measured in dollars, euros or another crypto asset. Exchange-rate movement affects the comparison. A dollar return should not be silently treated as the same result in the investor’s spending currency.
For a fictional example, an asset rises from 100 to 120 dollars, a 20% gain. If the local-currency value of one dollar changes from 1.00 to 0.90 during the same period, its local values become 100 and 108, an 8% gain before costs. The asset prices did not change in the example; the reporting unit did.
Record the conversion source and times. A BTC/USDT series and a dollar series require the quote-token treatment explained in the quote-currency guide. Do not assume every token-denominated balance is bank fiat.
Match instruments and observations
Use the same asset identity, market type and price field throughout. Spot, derivatives and company shares are different instruments. If a token migrated or changed units, read the source’s treatment before joining the series.
Match timezones and interval boundaries. A daily close on one schedule can differ from a rolling 24-hour field. A currently forming candle should not be treated as a completed historical close.
The snapshot worksheet preserves source, field and time. If definitions differ, show the limitation instead of forcing a precise ranking from incompatible inputs.
Calculate and normalise clearly
| Fictional asset | Start | End | Price change | Start indexed to 100 |
|---|---|---|---|---|
| A | $10 | $12 | +20% | 120 |
| B | $200 | $220 | +10% | 110 |
Indexing each start to 100 can make percentage movement easier to compare. It does not mean either asset originally cost 100 dollars. Label the axis as indexed performance rather than unit price.
The calculation is ending value divided by starting value, minus one, multiplied by 100. Preserve full precision internally. A rounded chart label should not replace the original data used to calculate the result.
Absolute price changes answer another question. Asset B rose by 20 dollars per unit while A rose by 2, but A had the larger percentage return in the invented example. Quantity held determines the actual money change for a person.
Include costs consistently
Compare both assets before costs or both under an explicit cost scenario. Do not subtract trading fees from one while leaving the other’s costs out. Funding, conversion, custody and withdrawal can affect a complete route.
A gross 10% price increase can produce a smaller net return when buying and selling charges apply. The return guide provides a worked fee convention. Use actual receipts for completed results rather than a generic percentage if the service charges differently.
Keep taxes and jurisdiction-specific reporting separate unless the model explicitly includes the applicable rules. An educational market ranking is not a tax calculation or a determination of the best instrument for a person.
Do not omit failed or unavailable observations
A comparison limited to assets that still trade can conceal assets that disappeared, lost access or lack usable data. State the selection universe and how missing cases are handled. A missing price is not automatically a zero return or an unchanged value.
If a new asset did not exist at the selected start, it cannot have the same full-period history without a different method. Mark the shorter record instead of inventing a baseline. A recent listing’s large percentage should not be ranked as though it covered an older asset’s entire interval.
Likewise, preserve data gaps and corrections. A provider’s revised historical series can change a result. Record the version or retrieval date where reproducibility matters.
Separate price rankings from portfolio experience
A person buying several times has different weights and dates from a one-unit chart. Deposits and withdrawals further affect performance measurement. Do not call an account balance increase a return without separating new money.
The average-cost tool can organise acquisition costs under its convention. Multiple sales and cash flows may need a more detailed analysis than a simple purchase average. Keep individual transaction records for that reason.
An unsold holding’s value is unrealised. A price comparison does not establish that the full quantity can be sold and withdrawn at the selected endpoint. Liquidity and service access remain separate questions.
Report the result with its limits
State the period, reporting currency, source, field, selection universe and included costs. Show the main figures and explain missing data. A small table can communicate those assumptions more clearly than a dramatic winner label.
The percentage guide explains changing denominators and compounded moves. Use it to check arithmetic, while keeping the future separate from the historical record.
A useful performance comparison can show which observation grew more under a stated method. It does not prove superior future returns, lower risk or suitability. Preserve that boundary even when one asset’s historical number is much larger.
Check the effect of an alternative window
A second predefined period can show how sensitive the conclusion is to the baseline. Report it beside the original result rather than selecting only the window that favours one asset. If a ranking reverses, explain that dependence. The comparison remains useful when its sensitivity is visible, while hiding the alternative would make a limited observation sound more general than it is.
Continue reading
English content and examples reviewed: 27 October 2026. Read the Arabic edition.
