Retail Forex Trading Strategies: Methods, Evidence and Risks

Candlestick chart with trend lines and currency symbols over a world map, titled “Retail Forex Trading Strategies.”

A forex strategy can win nine trades out of ten and still lose money. When comparing strategies for forex trading, the winning percentage tells only part of the story: a few large losses can erase many small gains, while trading costs can consume what remains.

The main approaches include following trends, trading ranges, entering on breakouts, responding to economic news, and seeking carry from interest-rate differences. Each relies on particular assumptions about the market. Understanding those assumptions helps explain both the opportunity and the conditions under which an approach can fail.

During our years working at forex brokers, we encountered clients running automated systems that opened sequences of positions, others concentrating on just a few economic releases, and traders mirroring orders across several accounts. Those observations gave us a close view of how people traded—and why a profitable episode could leave important questions about risk and repeatability unanswered.

This guide combines that brokerage perspective with research to explain how the main approaches work, where their weaknesses lie, and what a meaningful performance test should include. We’ll begin by separating four things that strategy lists often mix together: trading logic, holding periods, analytical tools, and position management.

What Makes a Forex Trading Strategy Complete?

“Day trading” tells you that positions are opened and closed within the same trading day. It leaves an important question unanswered: why enter the trade in the first place? A day trader might follow a trend, trade a range or respond to economic news. The holding period alone does not explain the trading logic.

To compare strategies for forex trading, separate four parts of the approach:

Trading strategies

Four parts of a trading approach

These four parts describe different aspects of a trading approach.

Parts of a trading approach, the questions they answer and examples.
PartQuestion it answersExamples
Trading logic

What behaviour or source of return does the approach seek to exploit?

Trends, mean reversion, breakouts, economic news, carry

Holding period

How long are positions normally held?

Scalping, day trading, swing trading, position trading

Analytical tools

What information helps identify and assess a trade?

Indicators, chart patterns, support and resistance, economic data

Order and position management

How are orders placed, positions sized and exposure changed?

Fixed position sizes, grid orders, martingale sizing, exposure limits

These elements can be combined in different ways. A breakout entry, for example, can form part of a trend-following strategy. A swing trader can trade either continuing trends or reversals.

From a trading idea to rules you can test

Consider a hypothetical trader who uses a moving average to identify a trend, buys when price breaks above a recent high, closes before the trading day ends and keeps position size fixed. That combines trading logic, a holding period, an analytical tool and a sizing rule.

It remains an outline. Which moving average? How is the recent high defined? What triggers an exit if the trade goes wrong? How much total exposure is allowed?

A strategy becomes meaningfully testable when those decisions are specified clearly enough to apply consistently. Entry conditions, exits and position management all affect the result.

This distinction matters when assessing indicators and automated systems. An indicator supplies information; its settings alone leave trading decisions unresolved. A grid distributes orders across price levels, while martingale sizing increases stakes after losses. Neither label establishes that the underlying trades have a profitable edge.

With those distinctions in place, we can compare the main approaches by the assumptions they make—and the conditions under which those assumptions can fail.

Compare the approaches

How forex trading approaches differ

Choose an approach to explore its core assumption, an illustrative rule and the conditions that can undermine it.

Trend followingPrice persistence

Follow a move already under way

Trend following seeks to participate in sustained upward or downward moves. A signal identifies a move that has begun; its continuation remains uncertain.

Core assumption

Some moves persist far enough to offset losing signals and trading costs.

Trend persistsMove reverses
Schematic price paths; no market data or forecast.
Illustrative rule

Buy when a 20-period moving average crosses above a 50-period average; exit on the reverse cross. Fix the chart interval and exposure limit before testing.

Where it can fail

Sideways markets produce repeated signals and reversals. A trend can also end soon after entry, giving back part of an earlier move.

Costs to check

Spreads, commissions and slippage; financing if held overnight. Frequent signal reversals increase turnover and cumulative costs.

Before you test

Record every signal, including failed ones. Check sensitivity to the chosen periods and performance on data that did not shape the rules.

Range tradingReversion within a range

Trade a return toward familiar prices

Range trading seeks rebounds between defined boundaries. It is one form of mean reversion, which more broadly targets a return toward a specified reference value.

Core assumption

The chosen boundaries or reference remain relevant long enough for price to revert.

Range holdsBoundary breaks
Schematic price paths; no market data or forecast.
Illustrative rule

Buy near a defined range’s lower boundary and exit near its midpoint. Specify how the range, entry tolerance and loss exit are calculated.

Where it can fail

A range becomes a sustained trend. A price that appears unusually low can keep falling, and a historical average can lose relevance.

Costs to check

Round-trip costs can absorb a small rebound. Entry and exit order types also affect the prices actually available to trade.

Before you test

Identify the range using only information available at the time. Include breakouts in the results and keep the loss rule fixed.

BreakoutsMovement beyond a boundary

Enter when price leaves a defined range

Breakout methods use a move beyond a specified high, low or boundary as an entry trigger. A breakout can also provide the entry for a trend-following strategy.

Core assumption

Movement beyond the boundary continues far enough to cover failed breaks and costs.

Move continuesBreak fails
Schematic price paths; no market data or forecast.
Illustrative rule

Buy after a bar closes above the previous 20 bars’ high; exit on a close back below that level. Fix the bar interval and position size.

Where it can fail

Price crosses the boundary and then returns inside it. Repeated failed breaks can accumulate losses before a sustained move appears.

Costs to check

Slippage around triggers, wider spreads during rapid moves and the cumulative costs of repeated entries and exits.

Before you test

Exclude the current bar when calculating the prior high. Model the next available fill after a signal, rather than assuming execution at the boundary.

News & macroRepricing expectations

Trade a change in the economic outlook

News and macro approaches assess economic releases, policy decisions and changing expectations. The surprise relative to expectations can matter more than the headline alone.

Core assumption

New information produces a repricing that the method can capture at executable prices.

Repricing persistsInitial move reverses
Schematic paths; a release can move price in either direction.
Illustrative rule

Trade only when a release differs from a recorded forecast by a defined amount. Specify the expected direction, entry delay and exit rule in advance.

Where it can fail

Details conflict with the headline, expectations were misread or the initial move reverses. The adjustment may occur before the order can be filled.

Costs to check

Spread widening, gaps and slippage around releases. A favourable chart move may offer little opportunity at the prices actually available.

Before you test

Use the forecast and first-released data available at the event, not later revisions. Include the delay and execution conditions of the intended trading setup.

CarryInterest-rate differences

Seek carry while accepting currency risk

A currency carry trade takes exposure to a higher-yielding currency funded by a lower-yielding one. Exchange-rate losses can outweigh the interest benefit.

Core assumption

The net carry is positive, and adverse currency moves do not erase it over the holding period.

Conceptual components. Other trading costs also affect the result.
Illustrative rule

Hold a higher-yielding currency against a lower-yielding one only when expected net carry is positive. Define an exposure limit and exit condition.

Where it can fail

The higher-yielding currency falls, the funding currency rises or rate changes reduce carry. A crowded unwind can amplify the currency move.

Costs to check

Actual rollover or financing terms, spreads and commissions. Retail broker funding rates can differ from the headline interest-rate gap.

Before you test

Separate financing credits or debits from exchange-rate gains and losses. Use the actual long or short funding terms for the instrument being traded.

These approaches can overlap. A breakout can be a trend entry, and news can trigger either a trend or a reversal. The sample rules illustrate trading ideas; they are not complete or validated strategies.

Sources for the concepts

These sources explain the underlying concepts. They do not establish the profitability of the illustrative rules above.

Retail Forex Strategies Through an Institutional Lens

A long lower wick shows that price closed above the period’s low. Saying “institutions were accumulating, so price will rise” adds two conclusions the candle itself cannot establish: who was trading and what happens next.

This is where observation, explanation and prediction can become confused. A chart pattern summarises recorded prices. Seeing a rally after that pattern does not establish what caused the rally, identify the participants involved, or demonstrate that the sequence will reliably repeat.

Institutional traders have different objectives

Institutional activity includes hedging existing currency exposure, obtaining funding, managing customer flows, rebalancing portfolios and speculating on exchange rates. Dealers can also match opposing customer trades internally, reducing their need to trade elsewhere. These activities have different objectives, which gives us reason to question explanations that assign a single institutional intention to a chart formation.

For example, buying a currency to reduce an existing exposure does not necessarily express a bullish forecast. Understanding the purpose of a transaction matters when interpreting its possible consequences.

A useful explanation must allow itself to be wrong

Consider an interpretation in which a price rise confirms “institutional buying,” while a fall is reinterpreted as institutions “collecting liquidity” before the expected rise. If every outcome preserves the explanation, the claim becomes unfalsifiable as stated.

A testable version defines the pattern, the expected outcome and the prediction horizon before examining the results. It records failures alongside successes and compares outcomes with a suitable baseline. Selected screenshots cannot establish how frequently a signal works.

Technical signals can be investigated rigorously. Research examining orders at a large foreign exchange dealing bank found stop-loss and take-profit orders clustered around round numbers, providing a mechanism for certain support and resistance effects. Extending that finding to another pattern—or turning it into a profitable strategy—requires additional evidence.

Institutional researchers face these difficulties too. Quantitative investment firm CFM has documented how selecting and refining strategies using historical results can produce backtests that overstate subsequent performance. Professional resources do not remove the risk of mistaking noise for an edge.

When assessing a strategy, separate three questions: What did we observe? What might explain it? What evidence supports using it to make a trading decision? A convincing answer to one does not automatically answer the others.

Trend Following: Capturing Sustained Price Moves

Trend following uses a defined signal to take positions in the direction of an existing move. Moving averages, breakouts and past returns can supply that signal. The underlying assumption is that some moves will persist long enough to make participation worthwhile.

Waiting for evidence of a trend creates a trade-off: entry comes after the move has begun, and an exit signal may arrive after price has reversed. Sideways markets can produce repeated false starts, often called whipsaws.

A Hypothetical Trade

Suppose a trend-following rule opens a long EUR/USD position at 1.1000. Price reaches 1.1200, then retreats until the rule triggers an exit at 1.1150.

The trade captures 150 pips before costs, having surrendered 50 pips from its peak. This illustrates why the exit rule matters: a profitable trade can involve giving back part of the move.

Across a series of trades, however, those gains must also cover failed signals, trading costs and any financing charges. One attractive trend on a chart tells us little about the complete strategy’s performance.

What the Research Supports

Moskowitz, Ooi and Pedersen’s 2012 study documented time series momentum across 58 futures and forward instruments, including currencies. Their benchmark strategy used the previous 12 months’ excess return to determine position direction, with a one-month holding period.

That provides evidence for a precisely defined approach. Applying the idea to an hourly retail forex chart requires a separate test of the chosen signal, exits, sizing and execution costs.

The practical challenge is enduring unsuccessful signals while remaining able to participate when a sustained move develops. Range trading begins with a different assumption: price will return toward a defined area.

Range Trading and Mean Reversion

Range trading seeks to profit from rebounds between defined support and resistance areas. A trader might buy near the lower boundary or sell near the upper boundary, expecting price to move back into the range. These boundaries are usually treated as zones rather than exact prices.

The related idea of mean reversion is broader: a deviation from a reference value is expected to narrow. That reference might be a historical average or a modelled relationship. Choosing it requires justification; calculating an average does not establish that price must return to it.

A Hypothetical Range Trade

Suppose EUR/USD has been moving between 1.1000 and 1.1100. An illustrative rule buys at 1.1010, targets the midpoint at 1.1050 and sets a loss exit at 1.0980.

Execution at those prices would produce either a 40-pip gain or a 30-pip loss, before costs. The example assumes the boundaries were identified before entry, using information available at that time.

The weakness appears when the range breaks and price continues moving away. The loss exit determines how the strategy responds when its rebound assumption fails. Spreads and other costs also reduce the benefit of relatively small price moves.

What the Evidence Supports

Carol Osler’s 2000 New York Fed study examined support and resistance levels supplied by six foreign exchange firms. It found evidence that the levels helped predict intraday trend interruptions, although predictive power varied across firms and currencies.

Those findings concern particular levels and price behaviour. Assessing a trading strategy requires testing its entries, exits and costs across both successful rebounds and failed ranges.

The same boundary that a range trader expects to hold can become a breakout trader’s entry trigger.

Breakout Trading: When Price Leaves a Range

Breakout trading seeks to capture movement beyond a predefined high, low or range boundary. The assumption is that price will travel far enough beyond that level to make the trade worthwhile.

The boundary and entry rule need to be clear before the move happens. Entering as soon as price crosses a level and waiting for a candle to close beyond it are different approaches. Each produces different signals, entry prices and missed opportunities.

The Trigger Price and the Fill Price

Take the EUR/USD range from our previous example: 1.1000 to 1.1100. A trader places a buy-stop order with a trigger at 1.1110, seeking to enter if price breaks above the range.

A standard stop order becomes a market order when triggered, so its execution price can be worse than the requested price.

Suppose the entry fills at 1.1120 during a rapid move. Price then falls back, and the exit fills at 1.1080. The loss is 40 pips, before separate fees. Calculating from the entry trigger would have suggested a 30-pip loss. That difference matters when testing a strategy whose apparent advantage is small.

Why a Breakout Can Accelerate or Fail

Research by Carol Osler, using customer orders from one large FX dealing bank, found that stop-loss orders clustered around round numbers and could intensify price trends. This provides a possible mechanism behind some breakouts; it does not establish that every crossed chart level offers a profitable opportunity.

A false breakout occurs when price crosses the chosen boundary and subsequently returns. Testing must include these failures, consistent exit rules and realistic execution assumptions.

Economic releases can supply the catalyst for a breakout. Trading those releases introduces another variable: how the announcement differs from what the market expected.

News and Macro Trading: Changes in Expectations

News traders focus on events such as employment reports, inflation releases and central bank decisions. Broader macro strategies may hold positions for weeks or months, building a view around economic developments and expected policy.

Both approaches need to distinguish what happens in the economy from what the market already anticipated.

Why the Headline Can Mislead

Suppose a U.S. employment report shows 200,000 additional jobs, while the median forecast was 250,000. Employment has grown, but the release falls 50,000 jobs short of expectations. Calling the headline “good news” misses that distinction.

A Federal Reserve study of EUR/USD and USD/JPY found that exchange rates responded quickly to the unexpected component of economic releases, with the estimated price adjustment generally occurring within a few minutes. Trading volume also increased following releases that matched expectations. An absence of surprise therefore did not mean an absence of activity.

The practical implication is that correctly interpreting an announcement may still leave little opportunity at an attainable entry price. Testing an immediate-reaction strategy requires the forecasts and release figures available at the time, together with realistic execution delays.

A Brokerage Example: Trading Selectively

During our years at forex brokers, we knew a client who concentrated on major economic releases and traded just two or three times a month. His approach illustrates how narrowly a strategy can be focused.

Assessing its repeatability would require the full trading record, including losing events, drawdowns and execution costs. A few profitable releases reveal little about how the approach handles a disappointing announcement or an unexpectedly poor fill.

News and macro strategies often involve expectations about interest rates. Carry trading puts the interest-rate difference itself at the centre of the approach.

Worked example · Carry trading

Carry Trading: Interest Income and Currency Risk

Carry trading seeks to earn the interest-rate difference between currencies. Its classic form involves borrowing in a lower-rate currency and investing in a higher-rate currency, leaving the exchange-rate exposure unhedged. The opportunity depends on currency movements allowing the investor to retain enough of that interest advantage.

When the Currency Move Outweighs the Interest

Suppose a hypothetical position earns $30 in net financing credits over a month. During the same period, an adverse exchange-rate move produces a $200 trading loss.

Net financing credits
+$30
Exchange-rate loss
−$200
Combined result
−$170

The combined result is a $170 loss, before entry and exit costs. Looking only at the overnight credits would miss the larger change in the position’s value.

Carry can also become vulnerable when many participants unwind similar positions. Banque de France describes how these exits can amplify currency movements, further eroding the gains accumulated from interest-rate differences.

What a Retail Trader Actually Receives

For a retail margin account, the relevant figure is the broker’s actual overnight financing rate for the position.

OANDA, for example, calculates its rates from liquidity providers’ short-dated currency swap rates, adjusted for an administration fee. It quotes separate rates for long and short positions, and those rates can change daily. Subtracting two central bank policy rates therefore does not establish the credit a client will receive.

A meaningful carry test combines financing credits or charges with currency gains and losses, transaction costs and the exposure carried through adverse moves.

The approaches covered so far describe potential sources of return. Grid orders and martingale sizing address a different part of trading: how positions are opened and exposure changes.

Sources for the concepts

The amounts in this worked example are hypothetical. Source links open in a new tab.

Grids and Martingale: How Exposure Builds

A grid places orders at predefined price intervals. Its direction, position sizes and exit rules depend on the particular system. Grids can use equal-sized orders or combine order placement with increasing sizes.

A classic martingale doubles the stake after a loss. Trading adaptations may instead add larger positions while earlier losing positions remain open. This combines a recovery objective with increasing exposure.

A Smaller Rebound, a Larger Position

Consider a hypothetical grid that adds a buy each time price falls another set interval. All five purchases remain open:

Position sizing

Fixed size vs doubling

Sizes of five successive purchases and total open size.
Sizing ruleSizes of successive purchasesTotal open size
Fixed size

1, 1, 1, 1, 1 units

5 units

Doubling each new purchase

1, 2, 4, 8, 16 units

31 units

Buying more at lower prices reduces the basket’s weighted average entry price. A smaller rebound can therefore bring it back to breakeven before costs.

That improvement comes with a larger position. After the fifth purchase, the doubling version has 6.2 times the exposure of the fixed-size version. Each further adverse price movement acts on that larger amount. Even the fixed-size grid accumulates exposure as purchases are added.

What Our EA Examples Show

Our original Expert Advisor examples illustrate the distinction: one sequence uses approximately equal-sized orders, while another increases the size of later entries. Positions are then closed together as a basket near breakeven.

A completed recovery shows how the exit mechanism worked on that occasion. Evaluating the system also requires the largest floating loss and exposure reached before the exit, together with cycles that failed to recover.

Closed Profits Can Conceal Open Losses

A history of profitable basket closures can coexist with a substantial loss on positions still open. Account equity incorporates unrealised gains and losses, and insufficient margin can trigger forced liquidation before a hoped-for rebound.

Assessment therefore needs the full equity path, limits on accumulated exposure and a defined exit for a failed recovery. The percentage of winning baskets alone leaves these questions unanswered.

Position management

See How a Grid Builds Exposure

Compare equal-sized and doubling purchases on the same hypothetical EUR/USD decline. Every purchase remains open: watch the position grow as its average entry moves.

First purchase: 0.10 lots; gap: 20 pips; 5 open purchases; further fall: 20 pips.

First buy 1.10000 Final buy 1.09200 Current price 1.09000

Fixed-size grid

Total open size

0.50lots

50,000 EUR · 5× the first buy

Floating P/L
−$300
Loss per further pip
$5.00
Rebound to breakeven
60.0 pips

Weighted average entry: 1.09600

Grid with doubling

Total open size

3.10lots

310,000 EUR · 31× the first buy

Floating P/L
−$1,140
Loss per further pip
$31.00
Rebound to breakeven
36.8 pips

Weighted average entry: 1.09368

P/L and price breakeven exclude spreads, commissions and financing. All dollar amounts are USD.

Every buy is assumed to fill and remain open. The example does not model margin limits or forced liquidation.

What changes with doubling

The doubling basket needs a rebound of 36.8 pips, compared with 60.0 pips for the fixed-size grid. It carries 6.2× the exposure and loses $31.00 per further adverse pip.

Open size after each purchase

Fixed sizeDoubling
Cumulative open size after each purchase (lots).
BuyFixed sizeDoubling
10.100.10
20.200.30
30.300.70
40.401.50
50.503.10
Cumulative open lots on a shared scale
Purchase number runs left to right. Both series use the same vertical scale. This is a calculated example, with no live market data.
Purchase sizes and entry levels

Scroll across to compare all columns.

All sizes are in lots; every purchase remains open.
BuyEntry priceFixed
new size
Fixed
total size
Doubling
new size
Doubling
total size
Assumptions and calculations

Model: EUR/USD buy positions in a USD account. The first entry is 1.10000. One standard lot represents 100,000 EUR; one pip is 0.0001 USD per EUR. Each later entry is one chosen interval lower. The extra fall creates no additional orders.

Execution: every scheduled purchase is assumed to fill at its stated price. No positions close. The example excludes bid/ask spreads, commissions, financing, slippage, margin requirements and forced liquidation. A real account may be unable to open or maintain the positions shown.

Sizing: the fixed grid repeats the starting size. The doubling grid uses 1, 2, 4, 8… times that size. This models adding to an open losing basket; classic martingale doubling after closed losses is a different sequence.

Average entry = sum of (EUR purchased × entry price) ÷ total EUR purchased
Floating P/L = sum of [EUR purchased × (current price − entry price)]
Loss per further pip = total EUR purchased × 0.0001
Rebound to price breakeven = (average entry − current price) ÷ 0.0001

Reading the numbers: the average entry and required rebound are rounded for display. Reaching that average entry makes the modeled price P/L zero; transaction and financing costs would still need to be covered.

Background: OANDA: profit and loss calculations; OANDA: margin and closeout mechanics.

Source links open in a new tab.

How to Evaluate a Forex Trading Strategy

A profitable backtest is a starting point. Its value depends on how the rules were chosen, which costs were included and how much exposure the account carried.

Calculate the Average Net Outcome

Win rate becomes meaningful when combined with the size of winning and losing trades. Using gains and losses measured before costs:

Average net outcome per trade = (win rate × average win) − (loss rate × average loss) − average cost per trade.

Return to the 90% win rate from our introduction. In a hypothetical ten-trade record, nine wins of $10 produce $90, while one $100 loss leaves a $10 deficit. If each trade also costs $1, the total loss becomes $20, averaging $2 lost per trade.

Deduct each cost once. Results already measured after costs should not have those same costs subtracted again. This average describes the record being examined; its reliability as a forecast still needs testing.

Test Rules on Data Reserved for Evaluation

Trying many indicator settings, entry filters and exit rules, then retaining the best combination, can produce an attractive backtest by fitting historical noise. CFM’s research illustrates how selecting and refining strategies can inflate performance measured on the development data.

A practical response is to record the variations tried and evaluate the chosen rules on data that played no part in their selection. If those results prompt further changes, that period becomes part of the development history. Further evaluation needs fresh data.

Measure Drawdowns and Shared Exposure

Track peak-to-trough declines in account equity, including unrealised losses, and how long recovery takes. Check whether profits depend on a handful of exceptional trades. The largest historical drawdown remains an observation, rather than a ceiling on future losses.

During our brokerage work, we saw clients mirror orders across several accounts. When assessing the evidence for their trading logic, those copies should be treated as exposure to one shared signal.

Choosing an approach to study also depends on the time, data and execution it requires.

Test the arithmetic

Can a 90% Win Rate Lose Money?

Nine $10 wins. One $100 loss. Add $1 per trade in costs. Explore how win rate, payoff size and costs change the average outcome.

Win rate: 90%; average win: $10.00; average loss: $100.00; average net cost per trade: $1.00. All amounts are USD.

Worked example shown below. Interactive controls appear when JavaScript is available.

Net average per trade

−$2.00

Negative after costs

Average before the costs entered here
−$1.00
Breakeven win rate
≈91.82%

Win rate for zero net outcome, holding average payoffs and costs fixed. Rounded where needed.

How the total is built

Win contribution

+$9.00

90% × $10.00

Loss contribution

−$10.00

10% × $100.00, subtracted

Cost adjustment

−$1.00

Subtract $1.00 per trade

Each row contributes to the average per trade. Bars share a scale: negative to the left of zero, positive to the right.

Frequent wins cannot cover the larger average loss in this example.

This average describes the supplied inputs. Assessing whether it will persist requires separate testing; the calculation does not measure uncertainty or drawdown.

Formula and assumptions

Let p be win rate as a decimal, W the average win, L the average loss as a positive amount, and C the average net cost across all trades.

Average net outcome = p × W − (1 − p) × L − C
Breakeven win rate = (L + C) ÷ (W + L)

This simplified model uses two outcomes: win or loss. Flat trades need separate handling. For a historical record, use the same set of trades and the same account currency for every input. Average cost may include financing charges minus credits and any execution costs that have not already been counted.

The breakeven rate holds payoffs and costs fixed. If the rate lies outside 0–100%, exact breakeven is unattainable with those inputs. If both average payoffs are zero, the net cost alone determines the outcome.

Inputs accept two decimal places; money inputs are limited to $1,000,000 in magnitude. Contributions retain fractional cents, so small differences remain visible. The calculator does not use sample size, trade order, open-position losses or changing position sizes.

Sources

CME Group: The Mathematics of Trading Success explains combining win frequency with the average size of wins and losses. This calculator adds an explicit average cost adjustment.

OANDA: Financing fees describes overnight financing charges and credits. Use the actual terms and amounts for the account being evaluated.

Source links open in a new tab.

Which Forex Strategy Should You Study First?

An approach that needs decisions at 10 a.m. is a poor fit if you can check the market only in the evening. Start with the time, data and execution conditions you can support.

Use these practical filters to narrow your research:

What you can supportApproach to investigateFirst practical check
Review prices at scheduled timesTrend or range rules using daily dataHow orders and exits work between reviews
Monitor a defined trading sessionIntraday range or breakout rulesSpreads, slippage and the timing of decisions
Track economic releases and forecastsNews or macro approachesWhether the test uses information available when each decision was made
Track financing and currency exposureCarryActual account-specific overnight charges or credits.

These filters establish practical fit. The evidence needed to support a strategy remains the same.

Turn the Idea Into Written Rules

For your chosen approach, record the signal, entry, exit, position sizing, maximum combined exposure and cost assumptions. Keep a log of decisions and departures from the rules. CME includes methodology, risk management and a trading log among the core components of a trade plan.

A strategy you can describe precisely gives you something you can test. “Buy a strong breakout” leaves room for hindsight. Specifying the boundary, trigger, order type and exit makes the decisions inspectable.

If the required data or execution records are missing, keep that limitation attached to the results. An attractive chart cannot fill the gap.

From a Trading Idea to a Testable Strategy

Return to the 90% win-rate example. Nine $10 wins could not cover one $100 loss, and costs deepened the deficit. Winning frequency becomes meaningful alongside the average size of gains and losses.

Trend, range, breakout, news and carry approaches each provide a starting point for research. Evaluating strategies for forex trading means examining the precise rules, costs and total exposure behind the results. Selecting the strongest historical performance can capture chance fluctuations, which makes testing on data reserved for evaluation essential.

Begin with one approach you can describe precisely and test under realistic conditions. Be as clear about what would invalidate the idea as you are about what would trigger a trade. That gives you a basis for questioning the results and deciding whether the strategy deserves further investigation.

Knowledge check

Test Your Forex Strategy Knowledge

Answer key: each question is followed by its correct answer and explanation.

Ten short scenarios. Choose one answer, check the explanation, then continue. You can open a hint before answering.

Interactive scoring appears when JavaScript is available. You can also answer the questions below and open their explanations.

  1. A trader follows a price trend, holds positions for several days and adds orders at preset price intervals. Which part of the strategy do those additional orders describe?

    1. Trade logic
    2. Holding period
    3. Chart input
    4. Position management
    Hint

    Separate the reason for taking a position from the rules that change the position after entry.

    Answer and explanations

    Answer: D. Position management

    A. Trade logic Following the price trend describes the trade logic. Adding orders describes how the position is managed.

    B. Holding period Several days describes the holding period. The additional-order rule concerns the position itself.

    C. Chart input A chart input supplies information, such as an indicator or price series. Adding orders is an action taken to manage exposure.

    D. Position management Adding orders changes the size and structure of the position. It is a position-management rule, distinct from the trend idea or holding period.

  2. Which statement is an observation that can be checked from the price chart alone?

    1. Banks bought to trigger traders’ stop orders.
    2. Price closed above the previous day’s high.
    3. The breakout will continue for two sessions.
    4. A hedge fund’s purchases caused the breakout.
    Hint

    Ask which statement reports a recorded price event without inferring a motive or a future outcome.

    Answer and explanations

    Answer: B. Price closed above the previous day’s high.

    A. Banks bought to trigger traders’ stop orders. The claimed motive needs evidence beyond the price chart. A price move alone does not identify who traded or why.

    B. Price closed above the previous day’s high. The recorded close and previous high can be compared on the chart. Explaining the cause or predicting the next move requires further evidence.

    C. The breakout will continue for two sessions. This statement predicts future prices. The current chart cannot establish that the predicted continuation will occur.

    D. A hedge fund’s purchases caused the breakout. This is a causal explanation involving a particular participant. The price chart alone cannot establish that participant’s trades or their effect.

  3. A hypothetical EUR/USD buy enters at 1.1010, targets 1.1050 and exits at 1.0980 if the setup fails. With one pip equal to 0.0001, what are the planned gain and loss distances? Ignore costs and slippage.

    1. 30-pip gain; 40-pip loss
    2. 40-pip gain; 30-pip loss
    3. 40-pip gain; 70-pip loss
    4. 400-pip gain; 300-pip loss
    Hint

    Measure each distance from the entry. The full distance between the target and failure exit answers a different question.

    Answer and explanations

    Answer: B. 40-pip gain; 30-pip loss

    A. 30-pip gain; 40-pip loss The distances are reversed: (1.1050 − 1.1010) ÷ 0.0001 = 40 pips of gain, and (1.1010 − 1.0980) ÷ 0.0001 = 30 pips of loss.

    B. 40-pip gain; 30-pip loss The target is 0.0040 above entry, or 40 pips. The failure exit is 0.0030 below entry, or 30 pips. These are planned distances under the stated assumptions.

    C. 40-pip gain; 70-pip loss The 70-pip distance runs from the failure exit to the target. Loss distance is measured from entry to the failure exit: 0.0030 ÷ 0.0001 = 30 pips.

    D. 400-pip gain; 300-pip loss Dividing the price differences by the stated pip size, 0.0001, gives 40 and 30 pips. The proposed distances are ten times too large.

  4. A EUR/USD buy-stop triggers at 1.1110, but the actual buy fills at 1.1120. The position later closes at 1.1080. With one pip equal to 0.0001, what is the realized price loss? Ignore other costs.

    1. 10-pip loss
    2. 30-pip loss
    3. 40-pip loss
    4. 50-pip loss
    Hint

    Use the actual entry fill and actual closing price. The trigger price is not the entry price in this example.

    Answer and explanations

    Answer: C. 40-pip loss

    A. 10-pip loss Ten pips measures entry slippage: (1.1120 − 1.1110) ÷ 0.0001. The full trade loss also includes the decline from the trigger level to the closing price.

    B. 30-pip loss Thirty pips uses the trigger as if it were the fill. The actual entry was 1.1120, so the realized price loss is (1.1120 − 1.1080) ÷ 0.0001 = 40 pips.

    C. 40-pip loss The actual fill minus the closing price is 0.0040. Dividing by 0.0001 gives a 40-pip loss. A stop trigger does not guarantee execution at that price.

    D. 50-pip loss The 40-pip loss from the actual fill already includes the effect of the 10-pip entry slippage. Adding that slippage again would count it twice.

  5. A hypothetical jobs release reports 200,000 new jobs. The consensus forecast was 250,000, and the previous release reported 180,000. What is the surprise relative to the forecast?

    1. +20,000 jobs
    2. +50,000 jobs
    3. −20,000 jobs
    4. −50,000 jobs
    Hint

    The forecast is the comparison point for a release surprise. The previous release answers a different question.

    Answer and explanations

    Answer: D. −50,000 jobs

    A. +20,000 jobs 200,000 − 180,000 = +20,000 compares the release with the previous number. The surprise uses the forecast: 200,000 − 250,000 = −50,000.

    B. +50,000 jobs The actual release is below the forecast, so the surprise is negative: 200,000 − 250,000 = −50,000.

    C. −20,000 jobs Comparing with the previous release would give +20,000, not −20,000. Comparing with the forecast gives the requested surprise of −50,000.

    D. −50,000 jobs The surprise is actual minus forecast: 200,000 − 250,000 = −50,000 jobs. This arithmetic alone does not determine the currency’s response.

  6. A hypothetical carry position earns $30 in net financing credits and loses $200 from the exchange-rate move. With no other costs, what is the combined outcome?

    1. $170 loss
    2. $170 gain
    3. $230 loss
    4. $230 gain
    Hint

    Financing credits and exchange-rate P/L are separate contributions. Combine them with their signs intact.

    Answer and explanations

    Answer: A. $170 loss

    A. $170 loss The combined outcome is +$30 − $200 = −$170. Positive financing income can coexist with a larger exchange-rate loss.

    B. $170 gain The currency loss exceeds the financing credit. The net amount is −$170, so reporting it as a gain reverses the sign.

    C. $230 loss Adding $30 to the loss treats a financing credit as a cost. The credit offsets part of the $200 loss, leaving a $170 loss.

    D. $230 gain This treats the exchange-rate loss as a gain. Keeping both signs gives +$30 − $200 = −$170.

  7. A hypothetical basket keeps five buys open, with sizes of 0.10, 0.20, 0.40, 0.80 and 1.60 lots. What is its total open size?

    1. 0.50 lots
    2. 3.10 lots
    3. 1.60 lots
    4. 3.20 lots
    Hint

    Every buy remains open. Total exposure includes the earlier buys as well as the final one.

    Answer and explanations

    Answer: B. 3.10 lots

    A. 0.50 lots 0.50 lots would be five equal buys of 0.10 lots. The stated basket increases each buy’s size, so its total is larger.

    B. 3.10 lots Adding all open buys gives 0.10 + 0.20 + 0.40 + 0.80 + 1.60 = 3.10 lots. This is 31 times the first buy’s size.

    C. 1.60 lots 1.60 lots is only the final buy. The earlier four buys remain open and bring the total to 3.10 lots.

    D. 3.20 lots 3.20 lots would be the size of a sixth doubled buy. The total of the five stated open buys is 3.10 lots.

  8. Ten hypothetical trades contain nine $10 wins and one $100 loss, all measured before costs. Each trade costs $1. What is the average net outcome per trade?

    1. $1 average loss
    2. $9 average gain
    3. $2 average loss
    4. $8 average gain
    Hint

    Add the ten trade outcomes, subtract the costs of all ten trades, then divide by ten.

    Answer and explanations

    Answer: C. $2 average loss

    A. $1 average loss The average is −$1 before costs: (9 × $10 − $100) ÷ 10. Subtracting the $1 cost per trade produces a −$2 net average.

    B. $9 average gain $9 is the win contribution per trade: 9 × $10 ÷ 10. It omits the $100 loss and the costs of all ten trades.

    C. $2 average loss The wins total $90, the loss is $100, and ten trades cost $10. Total net outcome is $90 − $100 − $10 = −$20; dividing by ten gives −$2 per trade.

    D. $8 average gain $8 subtracts the $1 average cost from the $9 win contribution but omits the $10-per-trade contribution from the losing trade. Including it gives −$2.

  9. You test chosen daily trend rules on a period reserved for evaluation. After seeing the results, you change the parameters. How should that period now be treated when assessing the revised rules?

    1. It has now helped develop the revised rules.
    2. It remains a fresh test of the revised rules.
    3. It proves the revised performance will persist.
    4. It is independent because the period is recent.
    Hint

    Trace whether information from that period influenced the revised parameters you retained.

    Answer and explanations

    Answer: A. It has now helped develop the revised rules.

    A. It has now helped develop the revised rules. The results influenced the revision, so that period is part of development history. A separate evaluation of the revised rules needs data that did not shape them.

    B. It remains a fresh test of the revised rules. The period is no longer fresh for the revised rules because its results influenced their design. Reusing it as untouched evaluation would overlook that dependence.

    C. It proves the revised performance will persist. Parameter changes can fit fluctuations in the period being examined. The revised historical result cannot establish that the same performance will persist.

    D. It is independent because the period is recent. Recency does not remove the influence of that period’s results on the revised rules. Independence from rule development is the relevant issue.

  10. Three accounts copy identical EUR/USD buys at the same time. What is the most accurate interpretation of their market exposure?

    1. They create three independent market return sources.
    2. Their combined exposure equals their average size.
    3. One copy counts; the other copies add no exposure.
    4. They share one market idea, and their sizes add.
    Hint

    Account separation does not change the pair, direction or timing of the positions. Consider what happens to all three when the same price moves.

    Answer and explanations

    Answer: D. They share one market idea, and their sizes add.

    A. They create three independent market return sources. All three positions depend on the same currency move in the same direction. Separate accounts do not make those copied market exposures independent.

    B. Their combined exposure equals their average size. Combined open size is the sum of the copied positions, not their average. Averaging would understate the total amount exposed to the currency move.

    C. One copy counts; the other copies add no exposure. Every open copy carries market exposure. Ignoring two copies would miss their additional gains or losses when the same price moves.

    D. They share one market idea, and their sizes add. The positions share the same market risk, and their open sizes add. Assess their combined exposure as well as any account-specific terms.

About this quiz and sources

The questions use this article’s definitions and hypothetical examples. Each has one answer; the explanation appears after checking. Restarting clears the current attempt.

CME Group: mathematical expectation — combining win frequency with average win and loss sizes.

OANDA: order types — order triggers, execution prices and slippage.

Federal Reserve: macroeconomic releases and FX trading — release surprises relative to expectations.

CFM: in-sample overfitting — how research choices can inflate historical performance.

Source links open in a new tab.

Sources & further reading

References

Research and documentation behind the article, with notes on the sections each source supports.

13 references · 2 topics

Research & market mechanicsReferences 01–07
  1. Reference 1. Global FX markets when hedging takes centre stage (opens in a new tab)

    Wenqian Huang, Ingomar Krohn & Vladyslav Sushko

    BIS Quarterly Review December 2025

    Institutional lens: hedging, funding, portfolio rebalancing and speculative activity help explain why participants trade.

  2. Reference 2. Time Series Momentum (opens in a new tab)

    Tobias J. Moskowitz, Yao Hua Ooi & Lasse H. Pedersen

    Research page hosted by AQR 2012

    Trend following: evidence from a specified futures and forwards study, including currencies. Applying its findings to a retail rule requires separate testing. Original study data (opens in a new tab).

  3. Reference 3. Support for Resistance: Technical Analysis and Intraday Exchange Rates (opens in a new tab)

    Carol L. Osler

    Federal Reserve Bank of New York · Economic Policy Review, 6(2) July 2000

    Range trading: an evaluation of support and resistance levels supplied by six firms, with findings that varied by currency and firm.

  4. Reference 4. Currency Orders and Exchange-Rate Dynamics: Explaining the Success of Technical Analysis (opens in a new tab)

    Carol L. Osler

    Federal Reserve Bank of New York · Staff Report 125 April 2001

    Breakouts and market mechanics: stop-loss and take-profit order clustering at one large FX dealing bank. The journal version appeared in 2003.

  5. Reference 5. The High-Frequency Effects of U.S. Macroeconomic Data Releases on Prices and Trading Activity in the Global Interdealer Foreign Exchange Market (opens in a new tab)

    Alain P. Chaboud, Sergey V. Chernenko, Edward Howorka, Raj S. Krishnasami Iyer, David Liu & Jonathan H. Wright

    Federal Reserve Board · International Finance Discussion Paper 823 November 2004

    News and macro trading: announcement surprises, rapid price responses and trading activity in historical interdealer EUR/USD and USD/JPY data.

  6. Reference 6. Carry trades and volatility risk (opens in a new tab)

    Xavier Denis, Tamaki Descombes & Henri de La Guéronnière

    Banque de France 11 February 2026

    Carry trading: how exchange-rate volatility and position unwinds can affect returns from interest-rate differences.

  7. Reference 7. In-sample overfitting: Avoiding the pitfalls in data mining (opens in a new tab)

    Capital Fund Management

    Technical note · PDF May 2016

    Strategy evaluation: simulations and the firm's experience illustrate selection bias, overfitting and the need for separate evaluation data.

Execution, costs & planningReferences 08–13
  1. Reference 8. The Mathematics of Trading Success (opens in a new tab)

    CME Group

    Trading Psychology · Education Undated

    Expectancy: combining win frequency with average wins and losses. Our worked example and calculator additionally account for trading costs.

  2. Reference 9. Building a Trade Plan (opens in a new tab)

    CME Group

    Education course Undated

    Choosing an approach: written objectives, methodology, risk management, strategy rules and a trading log.

  3. Reference 10. Understanding and using order types in leverage trading (opens in a new tab)

    OANDA

    Basic concepts · Education · BVI website Undated

    Range and breakout execution: how market, limit and stop orders work, including execution away from a stop's trigger price.

  4. Reference 11. Financing fees (opens in a new tab)

    OANDA Corporation

    Trading documentation · US website Undated

    Carry trading and costs: overnight financing charges or credits, buy/sell funding rates and administration fees.

  5. Reference 12. OANDA’s forex spreads and margins (opens in a new tab)

    OANDA Corporation

    Trading documentation · US website Undated

    Trading costs and grid exposure: changing spreads, margin requirements and the broker's margin closeout process.

  6. Reference 13. Calculating profit and loss on your trades (opens in a new tab)

    OANDA Corporation

    Trading documentation · US website Undated

    Worked trades and the exposure calculator: position-size and exchange-rate arithmetic in a USD-denominated account.

Author’s experience. The brokerage passages draw on our firsthand professional observations. Worked trades, calculator presets and quiz scenarios are hypothetical teaching examples.

Source scope. Broker documentation describes the named provider’s terms; details can vary by account and jurisdiction. “Undated” means the page does not show a publication date.

Sources checked: .

Do the Right Thing. Subscribe to Receive Our Updates!