Introduction
The safety of running an EA (automated trading) depends heavily on one thing: whether your orders are registered on the broker’s server.
Pending orders and SL/TP are typically stored on the broker’s server, so the broker can monitor and execute them even if your VPS or MT5 terminal goes down. In simple terms, you still have a minimum level of protection even when your terminal is offline.
By contrast, market entries and market exits rely on real-time communication at that exact moment. Higher latency or a brief loss of connection can quickly turn into real risk.
Comparison Table: Order Types × Server Registration × Latency Sensitivity (Key Terms Included)
| Order Type (JP / EN) |
Category | When to Use | Server-side? | Latency Sensitivity |
Key Notes |
|---|---|---|---|---|---|
| Market Order | Entry | Enter immediately at the current price | No | High | Strongly affected by connection quality and execution conditions |
| Market Close | Exit | Close immediately at the current price | No | High | You cannot send the close command while offline |
| Buy / Sell Stop | Entry | Enter on a breakout (opens a trade when price reaches a level) |
Yes | Low | Trigger is monitored server-side (fill depends on market conditions) |
| Buy / Sell Limit | Entry | Enter on a pullback (buy/sell at a better price) |
Yes | Low | Use Expiration to reduce the risk of “forgotten” pending orders |
| Stop Loss / SL | Protection | Limit losses automatically (closes when price hits the level) |
Yes | Low | Can slip during gaps or fast markets |
| Take Profit / TP | Protection | Lock in profits automatically (closes when price hits the level) |
Yes | Low | Fill price depends on liquidity and market conditions |
| Trailing Stop | Assist | Trail the stop as price moves (helps optimize exits) |
Usually client-side* |
Med–High | Often stops working if MT5/VPS goes offline |
* In standard MT4/MT5 behavior, a trailing stop runs on the client terminal (your MT5). Some brokers may offer “server-side trailing” as a separate service. Always confirm behavior on a demo or a small live account before relying on it.
Important: Latency (delay) and slippage are not the same thing. Even with server-side orders, you can still see slippage during fast markets, low liquidity, or gaps. See also: What is Slippage in Forex Trading.
Comparison: What Protects You When VPS/EA Stops? (Market-dependent vs Server-side Orders)
A Design That Needs Market Close
- If your VPS/EA goes offline, you can’t send a close command
- In fast moves, your position may be left open and unmanaged
- Your risk grows with downtime (until the system comes back)
- Client-side trailing stops often stop working too
Server-side SL/TP (+ Pending Orders)
- The broker server monitors SL/TP (they can still trigger while your VPS/EA is down)
- When price hits the trigger, the trade can close automatically—making it easier to cap losses
- For pending orders, using Expiration helps reduce “forgotten order” risk
- Treat trailing stops as exit optimization, not your main safety layer
Recommendation: As a rule, set SL/TP at the same time you open a position. For new entries, using server-side pending orders (Stop/Limit) often leads to a more stable and safer design.
Server-side vs Client-side: The 3 Risks You Must Design For (With Real-World Scenarios)
Accidents in EA trading often come down to where orders are monitored and executed, not the trading logic itself.
The core idea is simple: building around server-registered orders (Stop/Limit, SL/TP) usually makes your system more resilient to downtime and latency.
Below are the three key risk areas—latency, downtime, and fast markets—and what they mean in practice.
1) Latency — Where Market Orders Often Break Down
What happens: more bad fills and missed trades
Market orders depend on real-time communication, so higher latency makes it easier to get worse-than-expected fills and missed entries/exits.
Stop/Limit and SL/TP triggers are generally monitored server-side, so terminal latency tends to matter less at the “trigger” moment.
What to check: does your strategy still work when latency spikes?
Look for over-reliance on market entries/exits. Also, if your allowed slippage is too strict, you may see more rejections—meaning you can’t enter or exit when you need to.
Related: VPS Location for MT4/MT5 EAs: Latency Targets, Equinix NY4/LD4/TY3, and How to Choose
2) VPS/EA Downtime — Avoid the “Can’t Close” Failure
What happens: you can’t send the close command, so risk depends on downtime
When the VPS or EA stops, you often can’t send a Market Close command. If the market moves fast, an unmanaged position can turn into a large loss.
In that moment, server-side SL/TP is your last line of defense.
What to check: SL/TP at entry + Expiration for pending orders
Confirm your EA sets SL/TP at the time of entry (no “add it later” workflow). If you use pending orders, add Expiration so orders don’t get filled at an unintended time while your system is down.
3) Fast Markets & Gaps — Assume Slippage, Design the Limit
What happens: “exact price fills” are not guaranteed
During news spikes, thin liquidity, or weekend gaps, price can jump. Even pending orders and SL/TP can fill away from the requested level.
What to check: choose designs that survive slippage and costs
Instead of chasing “zero slippage,” build a position size, distance, and stop rule that can survive real execution. Very tight scalping systems tend to degrade when spreads widen or slippage increases.
Related:
Scalping EAs: Why They Often Fail on Live Accounts (Costs, Slippage, Execution)
Note: Separate “triggering” from “fill price”
SL/TP can trigger, but the actual fill price depends on liquidity and market conditions. Even so, using SL/TP makes it far easier to cap losses than running without a stop.
EA Order Design & Operations Checklist
This section is not about “making more profit.” It’s about preventing avoidable blowups.
Even a profitable EA can fail if execution rules and operations are weak. Use the checklist below as a solid defensive baseline.
1) Make Safety Mandatory at Entry
Set SL/TP at the Same Time You Enter (No “Add Later”)
Design your EA to place Stop Loss (SL) and Take Profit (TP) at entry.
Adding SL later can create “no-stop” positions during latency, execution delays, or a sudden shutdown. The goal is simple: zero trades without protection.
2) Remove “Forgotten Pending Order” Risk by Design
Use Expiration for Pending Orders
Stop/Limit pending orders can’t be canceled while your EA is down. Adding Expiration helps prevent accidental fills at the wrong time.
Before weekends or major events, pair Expiration with a quick audit of open orders.
3) Treat Trailing Stops as Optimization, Not Protection
Fixed SL/TP Is the Core Safety Layer
In many setups, MT5 trailing stops are client-side and may stop working during VPS downtime.
Use fixed SL/TP as your main safety layer, then use trailing stops only as an exit optimization tool.
4) Don’t Assume “Perfect” Market Conditions
Set Allowed Slippage and Stops Level Conservatively
Fast markets slip, brokers have constraints like Stops Level, and execution can be rejected.
If slippage is too strict, you may fail to enter/exit. If it’s too loose, you may get a worse fill than expected.
Build rules assuming rejections, jumps, and slippage can happen.
5) Use Monitoring So You Notice Failures Fast
Check VPS Status, Open Orders, and Error Signs
The most dangerous failure is not noticing your system stopped.
At minimum, monitor VPS uptime, open pending orders, and rising error rates. Alerts (uptime monitoring) help shorten “failure → awareness” time.
Summary: Build Around Server-side Orders to Reduce EA Accidents
Stable EA trading is not just about strategy logic. The foundation is whether your orders are registered on the broker’s server.
Market orders rely on real-time connectivity, so latency spikes and downtime can quickly create real damage.
Using server-side orders (Stop/Limit, SL/TP), adding Expiration to pending orders, and setting up monitoring can reduce the most common operational failures.
In practice, focus on a design that still protects the account even when the terminal goes offline.
FAQ
- Q. Do SL/TP still work if my EA or VPS stops?
- A. In most cases, yes. SL/TP are typically stored on the broker’s server and can be monitored and executed even if your terminal is offline. However, slippage can occur during fast markets or gaps.
- Q. Are Stop Orders and Stop Loss the same thing?
- A. No. A Stop Order is a pending entry. A Stop Loss is a protective exit for an open position.
- Q. Do pending orders guarantee no slippage?
- A. No. In gaps or thin liquidity, orders can fill away from the requested price. Triggering and fill price are separate.
- Q. Are trailing stops server-side?
- A. Standard MT4/MT5 trailing stops run on the client terminal. They usually require the terminal to stay online, unless your broker provides a special server-side trailing feature.
- Q. What if I want to cancel a pending order while my EA is down?
- A. Use Expiration to reduce “forgotten order” risk, and audit open orders after recovery.
- Q. What order types fit scalping best?
- A. Market orders are highly sensitive to latency. Many scalping approaches struggle when spreads widen or slippage increases. Server-triggered designs (Stop/Limit with SL/TP) tend to be more stable, depending on the strategy.
MQL5
Myfxbook
YouTube
X
GOGOJUN