Useful Statistics

Calculated the latest statistics on slippage in American futures since the beginning of the year. Let me remind you, that only the execution of stop orders is considered. Sometimes it happened to play market orders and limit orders, but I didn’t take it all into account — stop orders only. General impression, basically, developed — my goal was to determine how much to roughly expose the costs of a futures portfolio when testing on historical data — now I see that the generally accepted established standard is quite suitable, such as, на futures truth — 75 dollars per circle is quite normal. I generally, exhibiting 100 dollars per circle, since you still need to take into account the transitions from one month's contract to another, what often have to be done. I think, I will not continue to keep statistics, as, basically, and so everything is already clear.


System no. 1 It was 296 transactions, slippage $18.49
System # 2 was 156 transactions, slippage $28,59
System No. 3 was 64 transactions, slippage $10,44
System # 4 had 117 transactions, slippage $30,56

Total, in general, all systems were 633 transactions, slippage $22,40
It was not all round, but for a separate transaction. To count on a circle, must be multiplied by two 22,40 * 2 = $44,80

Also, counted for individual futures. Deals, truth, separately for each futures, Little, but the general impression can be made::

Intermediate Slippage Statistics.

Since the beginning of the year, I have been keeping statistics on slippage of futures instruments when opening / closing a position using a stop order. Everything on my sheet 37 futures.

  • Since the beginning of the year there have been 82 transactions.
  • IN 44 cases there was slippage.
  • IN 5 cases there was negative slippage, that is, in favor of the trader.
  • Average slippage per trade in dollars — 21,62
  • Average slippage in one trade in ticks — 2,48

This is not a circle, but for a transaction. That is, to find out in a circle, must be multiplied by 2. We get approximately 44 dollar or 5 ticks. If we add about 6 dollars or 1 tick on commission, then we get that when testing systems for a portfolio of futures, it is necessary to take into account the costs in one transaction 50 dollars per circle, or 6 ticks.

Most Outstanding Slips (>50 dollars) were:

  • Coffee — (281.25)  (56.25)  (56.25)
  • Gold (50)  (80)    
  • Palladium 175
  • Cotton 145
  • Lumber 66
  • Oat 50
  • Nicky 50
  • Platinum 65
  • Soybean oil 90

Slippage in futures

Calculated the real slippage for one contract on futures on the short-term system from those data, who stayed in Ninza, but stayed, unfortunately not all data.

Generally — 10 dollars per contract one way.

for each:

CL — $0 (on 13 traded contracts)
IT IS — $1 (12)
KC — $59 (6)
THE — $4 (23)
NG — $11 (11)
PL — $22 (12)
ZM — $7 (26)

Commission costs when trading futures.

Interesting, how much is it advisable to lay on commissions when testing American futures and, so called, slippage. Sometimes I meet an opinion, what if liquid ES futures are traded, then it is enough just to lay the commission 2 dollars per contract. Probably, mean that slippage on such a liquid futures makes no sense.
Hmm ….But the point is, what when testing, usually, no bid and ask, which are in real trade, there are only prices on the test, on which transactions took place. And the difference between bid and ask is always, least, at least one tick — for ES it is 12,5 dollars, actually. I doubt, that a trader will always be able to buy at a bid and sell at an ask — rather the opposite.
If the trader's counterarguments are that he only trades limit orders, then in this case those, the fattest" deals on tests, which opened only by touching a limit order and immediately reversed, let's go in the right direction, in real life will often not be performed, but those, which went right through and got a stop loss, in real life will always be performed. Therefore, the use of limit orders on tests sometimes, or rather, even always, lead to too optimistic results.
But okay, distracted — it's not about limit orders. So how much to pledge costs in tests on American futures? In particular, for ES, I'd, would not hesitate to lay 12,5 at the entrance, 12,5 to the exit, well, plus the broker's commission, eg 5 dollars per circle per contract. Total 30 dollars per transaction. And this, least.

At all , I have looked a lot of different commercial systems, or rather their tests. And somehow authors and independent experts, practically, are unanimous in their opinion and lay down costs approximately in the range of 25 to 75 dollars per circle. But these are modern, relatively new system builders. And if you dig deeper — 90-e years and the beginning of our century, then the standard was from about 75 to 150 dollars per circle.
In particular, independent firm, testing and comparing all kinds of systems for the futures market, FuturesTruthMagazine applies the following slip standards:

Commission/slippage used: $200/rt for ND, $100/rt for SP, $25/rt for e-minis, $75 for all other markets.

Who disagree? In particular, I would like to hear the opinion of the respected osmar92, since he trades fairly short-term trades in these futures — whether such costs are too critical for frequent intraday trades?

Why do I ask — just came up with a short-term system yesterday on a futures portfolio — the profit line goes up exactly, like a ruler, but the profit per trade is small — Total 98 dollars per contract…. And the test was without taking into account commissions and slippage :)

Scroll up