- Home
- Available brokers
- How we measure
How we measure each broker
Every number on a broker page is measured on Javlot’s own connected accounts, from real fills. Here is where each one comes from, without the jargon.
It starts with a reference account
At every broker we measure, Javlot runs its own account first.
A strategy signal fires on that reference account at the broker, then the very same signal is copied onto the connected client accounts. Almost every figure on a broker page is a comparison between the two: what the reference account got, and what the copied account got.
Because the comparison is always against Javlot’s own reference account at the same broker, the numbers describe how that broker behaved, not a gap between two unrelated places. None of it comes from a broker’s marketing. It comes from real positionPositionA single entry held on the broker account, with a direction, a size, and a current floating result.Click the word to learn more fills, the same ones you would have received on your account.
Copy speed
How quickly a position reaches your account after the reference account.
For every position, we take the moment the reference account was filled and the moment your copied position was filled, and subtract one from the other. That gap, in thousandths of a second, is the copy delay.
One position on its own tells you almost nothing, so we line up every measured position from fastest to slowest and take the middle one: the median. Half of the positions filled faster, half slower. The broker page shows that median, and the radar uses a stricter reading still, the speed that nineteen positions out of twenty beat.
Five measured positions, sorted from fastest to slowest:
| Position | Copy delay |
|---|---|
| 1 | 180 ms |
| 2 | 195 ms |
| 3 | 205 ms |
| 4 | 210 ms |
| 5 | 420 ms |
| Median, the middle value | 205 ms |
The slow 420 ms outlier does not drag the figure up, because the median only cares about the middle. That is the point of using it.
Timing consistency
Whether that delay holds steady, or jumps around.
A broker can be quick on average and still unpredictable. To catch that, we look at the spread of the delays: the fastest quarter of positions against the slowest quarter. When the two are close, the delay is steady. When they are far apart, it swings from one position to the next.
It is the difference between a broker that is always about 300 ms and one that is 120 ms one minute and 900 ms the next. Both can share the same median. Only the spread tells them apart.
Slippage, the price gap
Whether you got the same price as the reference account, going in and coming out.
You rarely enter at the exact price the reference account did. There is a tiny gap on the way in, and another on the way out. We measure both and add them together into one round turn, so a position is judged on the full trip and not just half of it.
The catch is that a pipPipThe smallest standard price increment on a currency pair, usually the fourth decimal.Click the word to learn more is a different amount of money on every instrument. One pip on gold is nothing like one pip on a euro pair, so pip counts cannot be pooled together. We first measure each gap as a fraction of the price instead of a fixed amount, a unit called basis points, on which gold, a currency pair and bitcoin all sit on the same ruler.
We then pool the gaps and show the result back in pips, which are easier to read. It is the same slippageSlippageThe gap between the price you asked for on an order and the price the broker actually filled it at.Click the word to learn more you feel on a market orderMarket OrderAn order that fills immediately at the best available price the broker offers right now.Click the word to learn more, measured against the reference account instead of the price on the screen.
A positive figure means a worse round turn than the reference account. A negative one means the copied position actually did better.
Commission
What the broker charged to open and close one standard lot.
Commission is read straight from the positions Javlot ran at the broker. We take what each position was charged, scale it to one standard lot sizeLot SizeThe unit a forex order is measured in. One standard lot equals 100,000 units of the base currency.Click the word to learn more, and take the typical value across recent positions, so the figure reflects the tariff in force today rather than one from two years ago.
Two cases get their own treatment. If a broker charges nothing per lot, we do not print a zero, we say “no commission”, because that broker earns from the spreadSpreadThe gap between the bid and ask price of an instrument, paid implicitly on every entry.Click the word to learn more instead, the mirror image of how an ecn brokerECN BrokerA broker that routes client orders directly into an electronic communication network of liquidity providers.Click the word to learn more charges commission and keeps the spread near zero.
And if we see more than one rate at the same broker, we do not blend them into a price nobody actually pays. We show each rate on its own.
The figure is always quoted per standard lot. A smaller position pays a proportional slice of it, here at a rate of 7.00 per standard lot:
| Position size | Commission charged |
|---|---|
| 1.00 lot (standard) | 7.00 |
| 0.10 lot (mini) | 0.70 |
| 0.01 lot (micro) | 0.07 |
Strategies available
How much of the Javlot catalogue can run at this broker.
This one is not a measurement, it is a count: how many of Javlot’s strategies are allowed at the broker, out of the whole catalogue, shown as “9 of 11”. It is a Javlot decision about where each strategy belongs, not a limit the broker imposes on you.
Available also does not mean already running there. It means the strategy is allowed at that broker, whether or not Javlot has started copying it yet.
When a figure is not shown yet
We would rather show nothing than show a number we do not trust.
Every figure needs a minimum of 100 measured positions behind it before it is published. Until then, the page shows how far along the count is, never a guess. Each figure sits in one of three states.
Javlot has no reference account at the broker, or has not run this measure yet. The figure stays empty rather than showing a zero.
Positions are being copied here, but fewer than 100 so far. The page shows the count, like “62 of 100”, so you can see it building.
At least 100 positions back the figure. Only now is an actual number published.
What we cannot measure yet
One cost is still out of reach today, held back by a technical limit we are working to lift.
Every figure on a broker page leaves out the spreadSpreadThe gap between the bid and ask price of an instrument, paid implicitly on every entry.Click the word to learn more, the gap between the buy price and the sell price that you cross on every entry. It is a real cost, and leaving it out is not a choice: measuring it cleanly needs price data our tools do not capture yet. We are working to change that, so a broker page can one day show the full cost of a position, spread included.
Until it is there, we never let a commission stand in for the whole bill. A low or zero commission does not, on its own, make a broker cheaper, because part of the cost still sits in the spread we cannot report yet. Wherever a commission appears, we say so plainly.
Terms used on this page
These figures describe what Javlot observed on real accounts over a period. They are not what a broker guarantees, and not what your own account will do. Past results do not guarantee future results.