Everything to Know Before Starting FXWhat 25x Really Means, Why People Lose, Taxes, and a Rule for Not Dying

Before starting FX, set two extreme expectations aside. Do not assume that leverage will change your life or that opening an account guarantees ruin. Start by examining what is traded and where losses actually expand.

This article is unnecessary for anyone who casually uses terms like pips, margin maintenance ratio, OCO, swap points, non-farm payrolls, and FOMC. It is written for people who are interested in FX but want to properly sort out the basics first: what exactly is being bought and sold, what 25x leverage concretely means, why the market stops on weekends, and what changes when London hours begin.

It will not be a mere glossary, though. There is no guaranteed winning method in FX. Not with AI, and not with automated trading. So the first thing to learn is not "how to make money" but "what am I actually doing."

1. The Conclusion: Design a Way Not to Get Knocked Out, Not a Way to Win

Three conclusions, stated up front.

• Leverage is not a device that multiplies your money. It is a device that amplifies the impact of price movement. The 25x maximum is not a recommendation — it is the line regulators drew because, left alone, retail traders break themselves.
• Most retail losses come not from misreading the market but from a missing design for money management and emotion. Deciding lot size first and hunting for a stop-loss distance afterwards is the classic failure.
• So the first thing to build is not a winning rule but a survival rule. How much you may lose per trade, how much loss ends your day, and the ceiling beyond which you will not add. Without those three, a high win rate still runs the account to zero eventually.

The rest of this article works through the specifics behind those three points.

2. What Are You Actually Buying and Selling?

What you trade in FX (foreign exchange margin trading) is neither a good nor a share of a company. It is the exchange ratio between two currencies.

If USD/JPY is at 160, it means one dollar equals 160 yen. Buying USD/JPY means buying dollars and simultaneously selling yen. Selling USD/JPY means selling dollars and buying yen. There is no special procedure equivalent to shorting a stock: buying and selling are almost the same operation. This is FX's defining structural feature — there is no constraint that you only profit in a rising market.

But being able to bet in either direction does not mean one of them will be right. The freedom to choose a direction always comes with exactly as much freedom to choose wrong. What actually matters is not your ability to predict a price ceiling or floor, but a single number: how large a position you hold relative to your margin. Everything else in this article is, in effect, a discussion of that one number.

For terminology, this much is enough to start.

Pay attention to the last two rows. Both the maintenance ratio and swap points are numbers that move on their own while you simply hold. From the moment you open a position, your situation keeps changing whether you act or not. Coming in with the mindset of holding a stock outright is where people stumble.

3. 25x Leverage Does Not Multiply Money — It Amplifies the Shaking

Retail FX in Japan is surprisingly young. The 1998 revision of the Foreign Exchange Act let individuals trade currencies directly, and through the 2000s brokers advertising 200x and 400x leverage were ordinary. Investor-protection rules then cut that back in stages: to 50x in 2010, and to the current maximum of 25x in 2011.

So 25x is not a level meaning "this much is safe." It is the line regulators drew after concluding that anything beyond it breaks retail traders. Reading a legal ceiling as a safety guideline is where beginners' mistakes begin.

The numbers make it concrete. With 1 million yen at 25x leverage, you hold a 25 million yen position. Your profit and loss then move against 25 million yen, not against 1 million.

USD/JPY moving a full yen in a day is not unusual. With an indicator release or an intervention involved, several yen in a day is entirely normal. In other words, the bottom row — the 4% that wipes out your capital — is not a once-in-decades catastrophe. Depending on timing, it is a distance the market can cover in a single day.

And "you may trade at up to 25x" is a completely different statement from "you should trade at 25x." Stack positions right up against the maintenance threshold and a small, entirely expected adverse move is enough to trigger forced liquidation. Your read on the market can be correct and you will still be removed before it plays out. Confusing those two while steadily increasing size is the most common way people take themselves out.

4. A Stop-Out Is Not a Safety Net

Most introductory articles claim you cannot lose more than your deposit because of stop-outs. That describes normal conditions. It is not a guarantee.

A stop-out force-closes your position when the maintenance ratio falls below the broker's threshold. That mechanism assumes a trade can actually be executed at that price. When the market gaps, the assumption itself breaks.

• Weekend gaps: if major news breaks over the weekend, Monday can open several yen away from Friday's close. No price existed in between, so a stop placed there is simply skipped over.
• Flash crashes: in thin liquidity, orders pile up in one direction and the price moves several yen within minutes. The early-morning plunge on 3 January 2019 is the textbook case — selling concentrated into a market emptied out by the New Year holiday.
• Currency interventions: the government and the BOJ move in trillions of yen. Direction, size, and timing are all unknowable in advance.

In those moments your fill lands far beyond the stop-loss you set. The result can be a negative balance exceeding the margin you deposited. "Stop-outs mean you cannot go into debt" is more accurately stated as "stop-outs usually contain it." Money management exists for the days that are not usual.

5. Why Markets Move (1): The Cast Changes by the Hour

The FX market runs 24 hours on weekdays, but the participants rotate through the day. Even in the same USD/JPY, morning and night are effectively different markets with different personalities.

If you feel the market suddenly got harder at night, that is not your imagination. An approach that works in a calm daytime range does not transfer intact into the London–New York overlap. Whether a method wins depends on the time of day as a precondition.

The market stops on weekends because it follows the business days of the banks and institutions that make it. World news does not stop while the market does — which is why Monday can reprice all of it at once.

6. Why Markets Move (2): Rate Differentials and the Gap Against Expectations

Why does the chart suddenly jump at 9:30 PM Japan time? Because that is when major US data — non-farm payrolls, CPI — is released (during daylight saving time).

The mistake beginners make most often here is assuming a good number means the price goes up. What the market reacts to is not the number itself but the gap against the prior consensus. If the market had priced in 200,000 new jobs and 180,000 arrives, that is still a decent number, and the dollar falls anyway. The expectation was already in the price before the release.

The reason that gap matters so much is that it changes the outlook for future interest rates. In USD/JPY, the dominant factor is the rate differential between Japan and the US. High US rates against low Japanese rates pull capital toward the higher yield, favoring a stronger dollar and weaker yen. A weak jobs report strengthens rate-cut expectations, the differential is expected to narrow, and USD/JPY falls.

Once you stop reading economic data as "good or bad" and start reading it as "which way does this move the probability of a hike or a cut," currency news suddenly becomes coherent. This is also the most genuinely interesting part of FX.

There is one more actor that moves price: the state. When the yen weakens excessively, the government and the BOJ intervene by selling dollars and buying yen on the order of trillions of yen. The state is not trying to profit from speculation; it is balancing domestic inflation, trade, financial stability, and diplomacy at the same time. That is why intervention is not something to bet on. Even if you read the direction correctly, you do not know the date or the size — and putting high leverage on that is gambling, not investing.

7. The Order Book Is Not a Map of the Future

FX has no single exchange the way equities do. It is a web of over-the-counter dealing, so a single consolidated book containing every order in the world cannot exist in principle.

The order books published by brokers such as OANDA show that broker's own customers' orders. Not the world. Mistaking one for the other means reading a local sample as the intent of the entire market.

Is it noise, then? Not quite. People like to place orders at legible places: round numbers such as 160.00, recent highs and lows, the previous day's close. As a visualization of where participants' attention is biased, the order book is genuinely useful. It is not a map that tells you the future; it is a map of what everyone is watching. That difference matters.

8. Most Losses Come From Your Side, Not the Market's

The failures that knock people out of FX look remarkably alike. Not because the market was unusually hard, but because human decision-making habits show up undiluted.

"It will come back if I wait a bit longer"

People feel the pain of a loss more strongly than the pleasure of an equal-sized gain — prospect theory, in behavioral economics. That asymmetry becomes the shape of your trading. Unrealized profit gets closed early because you want the relief; unrealized loss gets left alone because you do not want to make it real. The result is a structure that wins small and loses big.

The fix is procedural, not motivational. A stop-loss is a device your calm past self installs against your future emotional self. So you place it before you open the position. The moment you start wondering where to put it after you are already in, emotion has taken the wheel.

"I cannot sit still without doing something"

With a chart open, you want a position whether or not you have a reason. That is a response to boredom, not the output of analysis. Revenge trading right after a loss comes from the same circuit.

More trades means the spread — a fixed cost — accrues once per trade. If your method's expected value is unchanged, every additional trade makes you reliably worse off. The same applies to automated systems and AI: testing a thousand parameter combinations and adopting the best-performing one usually means you have excavated a coincidence, not an edge.

"I was right" and "that was a good trade" are different

Betting everything at full leverage and winning big produces a great result, but it is not a good trade. It is a lucky gamble. And a reckless success stays in memory as justification for doing it again. That memory is what eventually takes the whole account.

Conversely, if you cut the loss exactly as your rule specified, that trade was a good one even though it lost. Evaluate on whether you followed the procedure, not on the short-term result. Whether you can hold that standard is what separates the people who last from the people who do not.

Are institutions really hunting retail stops?

The claim that institutional players target retail stop-losses feels intuitively right but is structurally inaccurate. What large players need is enough liquidity to fill their own enormous orders. Liquidity naturally gathers at obvious highs and lows and round numbers — the very places retail stops sit. So it looks targeted. Nobody is looking at your account.

9. What the Data Says About Real Traders

Studies analyzing retail account data have repeatedly found that high leverage appears just as often among top performers as among the worst. In other words, leverage itself contains no power to win. It is purely an amplifier of whether the trade was right or wrong, and that shows up in the actual distribution of accounts.

FX is often described as a zero-sum market. Between speculators, one trader's gain is another's loss, but the market also contains commercial flows from exporters and importers, plus central banks. For an individual, spread, slippage, and swap are deducted on every round trip, so the account-level result behaves closer to negative-sum. That friction compounds with the number of trades.

Statistics about people who "made hundreds of millions in FX" deserve caution too. Multiple accounts, corporate accounts, and wealth generated by an entirely separate business and merely parked in an FX account all get mixed in. That successful traders exist and that their number can be counted accurately are two different claims.

The comparison of stocks versus FX as gambling also comes up often. USD/JPY essentially never moves 20% in a day, whereas an individual stock can go up tenfold in months. Raw price volatility is higher in equities. FX looks more dangerous because of leverage. Which means the gambling lives not in the instrument but in your own setting for how much risk to take.

10. Position Sizing: Derive Lot Size Backwards From Acceptable Loss

There is no universal answer to "how many pips before I cut." Twenty pips is right in some conditions; fifty is too shallow in others. Different volatility means a different appropriate distance.

The correct order is this.

1. Decide first how much you may lose on one trade (say 1–2% of capital)
2. Look at the chart and find the price at which your thesis is proven wrong — that is your stop
3. Work backwards from that distance to the lot size that keeps the loss within step 1

Reverse the order and fix the lot size first, and the stop gets placed wherever your capital can tolerate, which has nothing to do with the structure of the market. That is why it is more likely to keep getting taken out.

Say you have 1 million yen and accept 1% (10,000 yen) of risk per trade. If the stop sits 50 pips away (0.50 yen in USD/JPY), the size is 10,000 ÷ 0.50 = 20,000 units. If the distance is 100 pips, you drop to 10,000 units. Wider stop, smaller size. That alone keeps the impact of any single loss constant.

Once this clicks, the danger of averaging down becomes obvious. Adding to a losing position to lower your average entry is not inherently wrong. It is dangerous because an unplanned addition retroactively cancels the risk limit you set. A pre-designed scaled entry and one more lot added because you refuse to admit a loss look identical on the chart and are nothing alike.

Chasing swap points on high-yield currencies has the same trap built in. It looks like interest accruing daily just for holding, but it is not a savings deposit. High rates usually come attached to high inflation or political instability, and if the currency itself falls, months of accumulated swap income evaporate in a day.

11. Brokers, APIs, Automation, and What AI Actually Does

Choose a broker on fundamentals, not on campaigns or headline leverage. Registered with the Financial Services Agency; stable spreads; execution quality and stop-out thresholds stated explicitly. Those three are enough.

Unregistered offshore brokers advertising 500x leverage carry a triple risk: withdrawal problems, absence of legal protection, and the tax disadvantage covered below. There is basically no reason for a beginner to start there.

If you can program, brokers such as OANDA and Saxo offer APIs you can use to build your own trading tools. One warning matters above the rest: never embed an API key in browser JavaScript. The moment a key with trading permissions reaches a public surface, it is over. Keep it server-side, and implement a kill switch that halts trading at a daily maximum loss before you implement anything else. A warning as losses approach the limit protects capital; a confetti animation on a winning trade does not.

Automated trading (EAs) has existed for decades. The reason everyone is not rich has a clear explanation. Logic that wins in a trending market loses in a range. The more tightly you fit to historical data, the more the result is shaped by that specific period — overfitting. A backtest tells you how a hypothesis fared in the past; it guarantees nothing about the future.

AI sits on the same line. It is overwhelmingly strong at organizing news, generating backtest code, and computing indicators. But predicting tomorrow's USD/JPY is a different class of problem. Current language models fundamentally predict the continuation of text, and they do not hold news that has not happened yet. On top of that, any prediction effective enough to spread gets priced in and neutralized. Use AI as a tool for missing less, not as a tool for seeing the future.

12. The Practical Side of Taxes

Profits from a registered domestic broker in Japan are taxed separately from salary at a flat 20.315% (15% income tax, 0.315% special reconstruction income tax, 5% residence tax). The rate does not rise with size — but in exchange, FX losses cannot be offset against salary income.

The years you lose badly are precisely the years worth filing. FX losses carry forward for three years, but only if you file in the loss year and keep filing continuously thereafter. "I lost money so I do not need to file" is throwing away future tax relief.

"No filing required under 200,000 yen" is also widely misread. It is an exemption for salaried workers whose non-salary income stays under 200,000 yen, and it removes the income tax filing requirement only. The residence tax declaration remains. And if you file a return for any other reason — a medical expense deduction, for instance — FX profits must be included in it.

For expenses, what qualifies is what was directly necessary to earn the FX income. Claiming a general-purpose PC or an everyday AI subscription in full will not hold up. You need either evidence that it is dedicated to FX, or a defensible apportionment.

Finally, profits from unregistered offshore brokers fall outside that 20.315% separate taxation and are treated as miscellaneous income under aggregate taxation. Rates rise with total income, so the more you earn through high leverage, the smaller the share you keep. Run that calculation before concluding an offshore broker is advantageous.

(Tax rules change. Confirm against current guidance, and with a tax professional where it matters.)

13. A Template for the Survival Rule

Here is everything above, reduced to something you can write on paper before opening an account. Not a way to win — a way to still be standing tomorrow.

• Maximum loss per trade: cap it at 1–2% of capital, and write it as an amount ("20,000 yen"). Left as a percentage, you will stop computing it.
• Daily loss limit: reach it and you are done for the day. No attempts to win it back.
• Ceiling on effective leverage: being allowed 25x and using it are different things. Start around 3–5x and revisit only after you have seen how fast even that moves.
• Place the stop before entering: place it, then open. Never the reverse order.
• Decide which hours you do not touch: around major releases, the Monday gap, thin early mornings. Deciding what to avoid is a legitimate strategy.
• Keep a record: the reason for entry, where the stop was, the result, and whether you followed your rule. Recording rule-adherence in a separate column from win/loss is the whole point.

This rule will not raise your win rate. What it does is buy you the time to find one. As long as you are still in the market, you can keep learning.

14. Why FX Is Worth Doing Anyway

The Fed chair says one sentence and the rate jumps. Something happens in the Middle East and oil and currencies move together the next morning. The BOJ moves, and your mortgage rate and your account react in the same direction.

Once you start trading, economic news that used to be distant becomes your business. Behind the numbers you start to see states, corporations, and millions of intentions intersecting at once. There are not many mechanisms that let an individual observe the world's largest financial market from the same screen as everyone else.

In a place where astronomical capital and the sharpest minds collide continuously, there is no magic that wins easily. But you can treat it as an intellectual game: form a hypothesis, design the risk, verify the result. 25x is not an obligation, and you do not have to trade every day. The genuinely dangerous state is your trade size growing while your understanding of what you are doing does not.

The first skill to learn is not how to make a fortune. It is how not to be forced out. Learn that, and FX becomes an intellectual pastime you can keep for life.