Price analysis guide
Understanding Price Changes and Percentage Returns
A practical guide to measuring absolute and percentage price changes, defining observation windows, interpreting investment-equivalent examples and avoiding common errors.
What is a price change?
A price change describes the difference between two selected price observations. One is the starting value and the other is the ending value. Both must belong to a declared instrument and observation window, such as a session, day or multi-day period.
A positive change means the ending observation is higher than the starting one. A negative change means it is lower, and a zero change means the two selected values are equal. This describes endpoints; it does not summarize every movement between them.
What is an absolute price change?
Absolute change expresses the difference in the quote unit used by the market series. It is obtained by subtracting the starting price from the ending price. The sign preserves direction: a result above zero indicates an increase between the selected endpoints, while a result below zero indicates a decrease.
Absolute price change = ending price − starting price.
What is a percentage price change?
Percentage change expresses the endpoint difference relative to the starting observation. This makes changes from differently priced assets or periods easier to compare when the underlying definitions are compatible. The result remains dependent on the selected start.
Percentage change = ((ending price − starting price) ÷ starting price) × 100.
Reports should state their rounding convention. A displayed percentage may be rounded while the underlying calculation retains greater precision, so a manually reconstructed value can differ in its final decimal place.
How are price changes obtained reliably?
- Identify the instrument: state the asset pair, market or contract being measured.
- Define the window: record the start, end, timezone and exact cutoff timestamps.
- Select comparable fields: use a consistent price field for both observations.
- Validate the source: handle missing, duplicate, invalid or incomplete observations explicitly.
- Apply the declared method: preserve the formula, sign and rounding rules.
Reproducibility depends on the inputs as well as the formula. Two reports can both describe a 30-day change yet produce different results if their cutoffs, timezones, price fields or sources differ.
Why does the observation window matter?
The window determines which observations become the endpoints. A small change to either boundary can alter both the absolute and percentage result. A rolling period also changes as a new endpoint enters and the oldest one leaves.
Every published change should therefore travel with its timeframe and cutoff. A label such as “30-day change” is incomplete when the reader cannot determine which 30 days, timezone and observations were used.
What is an investment-equivalent example?
A report can apply a percentage change to a hypothetical starting amount to make the scale easier to understand. This is an arithmetic illustration, not evidence that an investment was executed or that every investor would have received that result.
Unless explicitly modeled, such an example does not represent fees, spread, slippage, taxes, financing, execution timing or risk controls. Those omissions should be stated so the illustration is not mistaken for an account-performance record.
What can price changes be useful for?
- Period summary: condense the difference between two declared endpoints.
- Compatible comparisons: compare percentage results across equally defined windows.
- Event detection: identify observations that cross a documented change threshold.
- Historical analysis: rank or group past windows after definitions and sample sizes are aligned.
- Model validation: supply a reproducible input or outcome for rules-based research.
These uses describe recorded data. A large positive or negative change does not prove a cause and does not guarantee continuation or reversal.
Limitations and common mistakes
- Endpoint dependence: the result can conceal a very different path between start and finish.
- Source dependence: separate markets can provide different prices for the same named asset.
- Field mismatch: comparing unlike price fields makes the result inconsistent.
- Incomplete periods: a live endpoint can change before the window closes.
- Cost omission: market price change is not automatically an investor's net return.
- Predictive overreach: historical change alone does not determine the next direction.
State the instrument, source, start and end timestamps, timezone, price field, formula, rounding rule, sample size and treatment of incomplete observations.
What are profit and loss?
Profit means a gain and loss means a negative result. In an accumulated-purchase simulation without sales, such as the linked study, the position’s value on a date is compared with contributions made up to that date. Value above contributions gives a positive result; value below them gives a negative result. Equal value and contributions represent break-even before omitted costs.
Total value is not profit: it includes contributed money. A valuation gain is also not a collected gain. The report simulates purchases and a closing valuation, with no real orders or documented sales; its profit is a model result, not a record of withdrawn cash.
How are profit and loss measured?
For this method, V is the value of accumulated BTC at the selected close and C is contributions accumulated by that time. The result retains its sign: positive means profit and negative means loss. With positive C, the percentage expresses that result relative to contributions.
Result = V − C. Percentage result = ((V − C) ÷ C) × 100.
The study supplies, without recalculation here, a final value of 1,353.26 USDT, contributions of 1,200.00 USDT and a result of +153.26 USDT (+12.77%). At its worst percentage close, it reports value of 278.63 USDT against contributions of 400.00 USDT and -30.34%. The bases differ because contributions increased along the path.
What can knowing both results help with?
It helps separate money added from changes in value, check whether the position exceeds its cost and see losses that a positive endpoint can hide. It also separates the combined result from each purchase’s result: in the linked sample of 12 purchases, nine finish profitable and three at a loss at the same close. They are not 12 independent strategy tests.
The analysis uses one window of 340 daily candles, valued at the October 6, 2026 close. It omits commissions, spread and slippage: it does not demonstrate an executable net gain. Its daily extremes compare value with contributions on each date, not a decline from a prior peak. Historical similarity does not guarantee the same future outcome.
Read the analysis of 12 monthly 100 USDT purchases, with ledger, source and report fingerprint.