Orders, stops & gap risk

BeginnerDuration ~10 min video + 25 min hands-onTools Your broker-spec.md from Lesson 1.1, A calculator

Why this lesson

Beginners treat a stop-loss as a magic force field. It is not. It is an order, subject to the same fills, delays, and gaps as any other order — and in the one situation you most need it, it can fail you completely. This lesson makes you distrust your stop in the right way, and shows the arithmetic that keeps a gap from ending your account.

Watch

Kevin Davey retells the single cleanest example of why a stop is not a guarantee: the 2015 Swiss Franc shock, on a system that had a stop on every trade.

Segment: ~1:25–3:50 — the SNB gapwatch full video

Watch for: Davey walks through the 15 January 2015 SNB de-peg: a roughly $21,000-per-contract move in about an hour that would have blown straight through any resting stop order, because a stop becomes a market order and there was simply no price near your stop to fill at. That is 'a resting stop is not a stop' during a gap. The rest of the video is his mean-reversion system and a workshop plug — skip it.

The explainer

Four order types cover almost everything. A market order says “fill me now at the best available price” — fast, but you accept whatever price exists. A limit order says “fill me only at this price or better” — you control price but not whether you fill at all. A stop order says “when price reaches this trigger, send a market order” — used to get out (a stop-loss) or to enter on a breakout. A stop-limit adds a limit price to the stop, which sounds safer but has a nasty edge: in a fast move price can leap past your limit and leave you unfilled and unprotected. For an exit stop, that is the opposite of what you want.

The critical, under-taught fact hides inside the stop order: when the trigger is touched, the stop becomes a market order. It does not fill at your stop price. It fills at the next price the market offers, which in calm conditions is a fraction of a pip away and in violent conditions can be very far away. That gap between your intended and actual fill is slippage, and it is not random noise — it leans against you on stops. FXCM’s published execution statistics show 57% of stop and stop-entry orders received negative slippage (FXCM slippage statistics); around news events, 10–20 pips of slippage is normal and 50+ occurs (slippage guide). We treat slippage lightly here and cost it properly in Lesson 1.3 — for now just hold the idea that a stop’s price is a wish, not a promise.

Now the distinction that separates hobbyists from operators: a broker-side stop versus a software stop. A broker-side stop lives on the broker’s server; it fires even if your laptop is asleep, your Wi-Fi dropped, or your Python script crashed. A software stop lives on your machine — your code watches the price and sends the exit order only when the level is hit. The problem is obvious once stated: if your machine, connection, or code fails at the wrong moment, the software stop does not exist. There is nothing between your open position and the market. Orphaned positions with no server-side protection are a classic bot-account killer, which is why the rule from day one is: the stop that protects you is the one sitting on the broker’s server. You’ll enforce this in Level 3, but you adopt the belief now.

Even a perfect broker-side stop has a hole, and it’s the same hole Davey described: the weekend gap. FX and CFD markets close from Friday evening to Sunday evening. News breaks while the market is shut. When it reopens, price does not walk from Friday’s close to the new level — it jumps. Your stop, triggered at the open, becomes a market order and fills at the first available Sunday price, however far below your stop it sits (gap risk explained). The tail case is the Swiss one: on 15 January 2015 the SNB abandoned its EUR/CHF floor and the pair gapped roughly 40% straight through every stop. It produced about $225 million of negative client balances at FXCM and bankrupted Alpari UK (event account, FXCM insolvency coverage). Stops did not fail because they were badly placed. They failed because there was no price near them to fill at.

There is exactly one order that caps this: a guaranteed stop-loss order (GSLO), a paid product where the broker contractually fills you at your stop price no matter the gap. You pay a premium for it, and it is the only thing that truly bounds gap risk. For everything else, the defense is not an order type — it is size. If you assume a stop might fill much worse than its level, or not protect you at all through a gap, then the only variable you actually control is how large the position is when that happens. Which is the exercise.

Do it

You’re going to size a position so a plausible weekend gap can’t wreck you, and see what a guaranteed stop is really worth.

1. From broker-spec.md, take EUR/USD: pip value ≈ $10 per standard lot ($1 per mini lot). Assume an account of $5,000 and a rule of risking 1% ($50) per trade.
2. Normal case. You place a stop 25 pips away. Position size = risk ÷ (stop distance × pip-value-per-lot). In mini lots: $50 ÷ (25 × $1) = 2 mini lots. Write that down.
3. Gap case. Now assume a weekend gap blows the price 150 pips past your stop before it fills (modest next to the SNB’s thousands of pips). With those same 2 mini lots, your realized loss = 150 × $1 × 2 = $300 — that’s 6% of the account from one trade you thought risked 1%.
4. Ask the real question: what position size keeps the gap loss (150 pips) to 1%? $50 ÷ (150 × $1) = about 0.33 mini lots. Notice how much smaller “gap-safe” size is than “stop-safe” size.
5. Write one sentence for your notes: “My stop distance sizes my normal risk; the gap I can’t rule out sizes my real risk.” Then note when a GSLO (paid) is worth it — typically for positions held over weekends or through scheduled high-impact events.

Keep this beside broker-spec.md. Lesson 1.4 turns “risk per trade” into the R-multiple.

Terms introduced

Terms introduced in this lesson

• market order
• limit order
• stop order
• stop-limit order
• broker-side vs software stop
• slippage
• weekend gap
• guaranteed stop (GSLO)

Each term links to the glossary, where it is defined by the way you just used it.

Check

Check yourself

1. A stop order, once its trigger price is touched, becomes a:

   limit order at the trigger priceguaranteed fill at the trigger pricemarket order that fills at the next available pricecancelled order

2. A "software stop" (a stop your script holds and only sends when hit) is dangerous because:

   it always fills worse than a broker-side stopbrokers charge extra for itif your machine, connection, or code fails, the stop simply does not existit counts as a limit order

3. Over a weekend gap, a normal stop protects you:

   fully, because brokers freeze prices over weekendsonly if you also set a limit ordernot at your price — it fills at the first traded price on Sunday open, however far awayexactly at your stop price

4. The only order type that caps your fill price through a gap (for a fee) is a:

   guaranteed stop-loss order (GSLO)trailing stopstop-limit ordermarket order

5. On 15 January 2015 the Swiss National Bank de-peg caused EUR/CHF to:

   rise smoothly with no gapsstop trading for a monthgap roughly 40% through every resting stop, producing negative client balances and bankrupting a major brokerdrift down 2% over a week

Check answers

You can move on when you can… explain why a software stop is not a stop, and size a position for a weekend gap.

Mark this lesson complete

Go deeper

• BabyPips — Types of Forex Orders: the clean free reference for market, limit, stop, and stop-limit mechanics.
• tastyfx on gap risk and the LeapRate SNB post-mortem: read both, then re-price your worst case.
• The book that owns this topic: Ernie Chan’s Quantitative Trading (ch. 5, execution systems) is the cleanest short treatment of how orders actually behave in a live system — the bridge from this lesson to Level 3.