Methodology & Data Sources
How the technical indicator readings on PriceFeed Pro are produced, and what they are not.
In short
- The readings are produced by a fixed, rule-based algorithm from historical prices. No human selects, edits or reviews any reading.
- The algorithm is deterministic: the same price history always gives the same readings, and every visitor and every API customer sees the same output. Nothing is personalised.
- A reading such as “Positive” is a count of how many indicators currently sit on one side of their rule. It is not a forecast, a recommendation or a trading signal.
- We have measured the readings against what price did afterwards and found no directional edge: they do not predict the next move better than chance.
Who operates PriceFeed Pro
PriceFeed Pro is operated by Omar Abbas, an individual based in Amman, Jordan; it is not a registered company. It provides market data and automatically generated technical-indicator readings, and does not provide personal investment advice. For legal and compliance enquiries, see the contact details in the Terms of Service.
Data sources
PriceFeed Pro is multi-broker. Each price source is a broker's own live trading server, read through a MetaTrader 5 terminal connected to that server. Each source has a source identifier, for example ORBEX-live-1, and every analysis response names the source it was computed from in its source_id field.
A broker that shows PriceFeed Pro readings on its own pages receives readings computed from its own live server feed, so the prices behind the readings are that broker's quoted prices. The public analysis pages on pricefeedpro.com currently use ORBEX-live-1.
Prices are broker bid prices. Indicators are computed on completed bars only: a bar that is still forming is never used. The pages show when the data was last updated and flag readings whose data is delayed.
Indicators and how each is read
Each timeframe (1m, 5m, 15m, 30m, 1h, 4h, 1d, 1w and 1mo) is read by the same two groups of indicators. A timeframe needs at least 30 completed bars (60 for monthly) before it is read at all; until then it is shown as having insufficient history. An indicator that cannot yet be computed is left out of the count rather than counted as neutral.
14 oscillators
| Indicator | Reading rule |
|---|---|
| RSI(14) | Positive when at or below 30 and rising; negative when at or above 70 and falling; otherwise neutral |
| Stochastic %K(14) | Positive at or below 20 and rising; negative at or above 80 and falling |
| CCI(20) | Positive at or below −100 and rising; negative at or above +100 and falling |
| Williams %R(14) | Positive at or below −80 and rising; negative at or above −20 and falling |
| ROC(10) | Sign of the 10-bar rate of change |
| MACD(12,26,9) | Sign of the MACD histogram |
| ADX/DMI(14) | When ADX is above 20: positive if +DI > −DI, negative if −DI > +DI; otherwise neutral |
| Bollinger(20,2) | Positive at or below the lower band; negative at or above the upper band |
| Aroon(14) | Positive if Aroon Up > Aroon Down; negative if the reverse |
| TRIX(9) | Sign of TRIX |
| Vortex(14) | Positive if VI+ > VI−; negative if the reverse |
| Parabolic SAR (0.02, 0.2) | Positive while price is above the SAR; negative while below |
| Supertrend(10,3) | Positive in an up-state; negative in a down-state |
| Coppock (14, 11, 10) | Sign of the Coppock curve |
The four bounded oscillators (RSI, Stochastic, CCI, Williams %R) only read positive or negative when they are at an extreme and turning back; an extreme that keeps extending reads neutral.
6 moving averages
SMA and EMA at 20, 50 and 200 periods. Each reads positive when price is above the average and negative when below. On timeframes without 200 completed bars, the 200-period averages are not counted.
How a timeframe's label is tallied
For each group, and for both groups combined, we count positive (buy), negative (sell) and neutral readings and compute the net share:
net share = (positive − negative) ÷ (positive + negative + neutral)
- net share ≥ 0.5 → Strongly positive
- 0.1 < net share < 0.5 → Positive
- −0.1 ≤ net share ≤ 0.1 → Neutral
- −0.5 < net share < −0.1 → Negative
- net share ≤ −0.5 → Strongly negative
“Strongly” means more of the indicators sit on one side; it is not a measure of confidence and does not mean a larger move is expected. In the API these five bands appear as STRONG_BUY, BUY, NEUTRAL, SELL and STRONG_SELL; the public pages show them as the words above.
Across timeframes
The combined reading of each timeframe counts +1 (positive), −1 (negative) or 0 (neutral), multiplied by a fixed weight. The cross-timeframe reading is positive or negative only when the weighted net is at least 60% of the total weight read; otherwise it is mixed. The same price history always gives the same answer, so a value near 60% can switch between mixed and directional as single timeframes change.
| Timeframe | Weight |
|---|---|
| 1mo, 1w | 3.0 |
| 1d | 2.5 |
| 4h | 1.5 |
| 1h | 1.0 |
| 30m | 0.65 |
| 15m | 0.5 |
| 5m | 0.35 |
| 1m | 0.25 |
Chart-pattern detectors
Separately, seven chart-pattern detectors check each timeframe for specific configurations (for example a moving-average stack or a MACD cross). They are usually all silent. The pages report how many are present; the detectors' internal score is not shown, because it is not calibrated as a probability.
Pivot points and support/resistance
Pivot levels are arithmetic on the previous completed day, week or month, where H, L and C are that period's high, low and close.
R1 = 2·PP − L · S1 = 2·PP − H
R2 = PP + (H − L) · S2 = PP − (H − L)
R3 = H + 2·(PP − L) · S3 = L − 2·(H − PP)
R1/S1 = C ± (H − L)·1.1/12
R2/S2 = C ± (H − L)·1.1/6
R3/S3 = C ± (H − L)·1.1/4
R1 = 2·PP − L · S1 = 2·PP − H
R2 = PP + (H − L) · S2 = PP − (H − L)
R1/S1 = PP ± 0.382·(H − L)
R2/S2 = PP ± 0.618·(H − L)
R3/S3 = PP ± 1.000·(H − L)
“Nearest support” and “nearest resistance” are the closest daily floor (classic) pivot levels below and above the current price. They are reference levels, not price targets.
What the readings are not
- Not a forecast. We record each reading and the price that followed. Measured that way, the readings show no directional edge: they are right about the next move roughly as often as they are wrong.Supporting detail: a replay test of 523 daily snapshots (22 June – 29 September 2026; XAUUSD, XAGUSD, NAS100, US30, EURUSD, GBPUSD, USDJPY and BTCUSD), each scored against the next daily bar, found directional readings right 44–47% of the time. For comparison, “always up” scored 50.7% and “same as yesterday” 46.4%. Higher scores were not more accurate. The replays were rebuilt from stored bars rather than from what was published live, and each rate is uncertain by about ±12 percentage points.
- Not a recommendation. The output is general information, identical for all users, and takes no account of anyone's objectives, finances or circumstances.
- Not advice from a person. No analyst writes, selects or adjusts any reading.
Trading leveraged products involves a high risk of loss. PriceFeed Pro accepts no liability for reliance on this information.