Every StaxInvesting article tagged risk management · 120 posts.
50 articles
Ahead of an earnings print, implied volatility inflates the premium; after it, the volatility crushes, and a long-premium position can lose even when the underlying moves the right way. How entry timing, price filters, structure, and sizing bound that failure mode, and the honest limit of each.
The two words describe opposite ends of the same event, and confusing them obscures the fact that only one side has a choice — the seller finds out afterward.
In options, long and short do not mean bullish and bearish. They mean which side of the contract you are on — and that determines your risk profile more than whether it is a call or a put.
Implied volatility explains why an option can lose money while the underlying moves in your favour — which makes it the concept most worth understanding early.
A scheduled catalyst lands at a known time, which turns the decision to trade the window into a configuration choice rather than a reaction. What schedule control governs, what it misses, and why it has to be paired to be complete.
A dispersed tape can leave the index calm while individual names move sharply in both directions, and a flat index offers no protection against concentration you did not price. How symbol filters and concurrency caps bound single-name risk, and the honest limit of each.
A low VIX ahead of a binary event does not mean the event is low-risk: it means options are pricing a small move against a catalyst that can produce a large one. How sizing, defined-risk structures, and loss limits bound that gap, and the honest limit of each.
A scheduled macro print widens the distribution of intraday outcomes, and risk settings tuned for an average day behave differently inside it. A mechanism-by-mechanism look at bounding automated exposure on a high-variance session â and the honest limit of each control.
Under intraday trailing, an unrealised high you never converted still raises your floor — which means traders fail while their realised results are positive. It is the rule most often misread, and the misreading happens during a breach.
Slippage is the reason a strategy can be profitable in a backtest and unprofitable in an account — and it is largest in exactly the conditions that produce a strategy's biggest moves, which is when a fixed assumption is most wrong.
Return describes where a strategy ended up. Drawdown describes what it did on the way — and that is the number that determines whether you are still following it at the end.
The detail most content glosses over is that these accounts are typically simulated and the firm's revenue comes substantially from evaluation fees. That is a structural fact, not an accusation — and it changes how a pass rate should be read.
The signal logic usually transfers. Everything around it does not — and the parts that differ are the parts that fail expensively, starting with a sizing routine that computes risk from the wrong number entirely.
Calling both of these leverage obscures more than it explains. The question is not which has more — it is which failure mode you are accepting: loss by magnitude, or loss by expiry.
Nearly-continuous access sounds like an unambiguous advantage. It is more accurately a different shape of market, with hours that behave nothing alike — and a day boundary that breaks daily counters written for equities.
Options traders arrive with the wrong mental model. An options buyer pays a premium and owns something; a futures trader posts collateral and owes performance — and the loss is not capped by what was posted.
This is the most useful fact for a smaller account entering futures, and it has no options equivalent — you cannot buy a tenth of an option contract. But ten micros cost ten commissions for identical exposure.
Options traders arrive at futures with no vocabulary for this. There is no premium and no strike — there is a specification sheet, and the three numbers on it determine your dollar risk per contract.
Futures automation shares infrastructure with options automation and almost none of its vocabulary. This is the map — contract specs, rollover, posted margin, and a session that barely closes.
Platform minimums answer what you can deposit, not what you need. If a provider's typical position costs more than a twentieth of your account, you cannot follow them correctly — and following incorrectly is worse than not following at all.
Understanding a marketplace as a business explains most of what is confusing about how strategies get presented — starting with why the default sort surfaces whoever took the most risk and has not yet been punished for it.
These get compared as competing products. They are structurally different arrangements — and copying protects you from misappropriation while doing nothing about market risk, which is where almost all the money in this category is actually lost.
Skipping trades is not random. It correlates with how the last one went, so you skip after losses and participate after wins — which means you are no longer running the strategy you subscribed to, but a filtered version chosen by your emotional state.
CME Globex runs roughly 23 hours a day, which means signals arrive while you are asleep and manual copying cannot participate in a large share of the session. It also means the same trade fills very differently at 10 AM than at 3 AM.
The single most reliable filter is whether losing periods are visible. An operation that only shows winners is not showing you a track record, it is showing you marketing. Written from the perspective of having lost money to exactly that.
The regulatory picture is genuinely unsettled in places, and content presenting it as simple is usually selling something. A sitting CFTC Commissioner has dissented publicly over where the line between software and advice sits.
The honest version of this question is arithmetic before it is aspiration. If the return rate your capital needs to produce your target income sounds impressive, that is a warning rather than a plan.
Evaluating a record well is mostly a matter of asking what the presentation is designed to obscure. Start with whether losing trades are shown at all — if they are not, there is nothing to evaluate.
Most followers think they are subscribing to a list of trades. They are subscribing to a risk posture, and the trades are how it expresses itself. You inherit the drawdown without inheriting the experience that makes it tolerable.
A leader can show a win rate above 57 percent and still hand followers losses, because win rate describes frequency rather than profitability. The gap between a headline record and a follower's actual result is structural.
Choosing a signal provider is the entire decision in copy trading; everything else is implementation. What matters is not their headline return but whether what they provide can be inspected or only inferred.
Social trading is usually presented as the friendly, educational end of the category. The evidence is more complicated: studies found socially influenced trades underperformed the same investors' independent ones.
Most beginner guides in this category are written by platforms that earn money when you participate. This one assumes you might reasonably decide not to — and starts with the four things that determine whether it goes well.
Five positions in correlated instruments is closer to one large position than five small ones. The setting counts positions; risk depends on exposure. Those are the same thing only when correlation is low — and correlation rises exactly when it matters.
This is the one risk setting where the honest answer may be that you should not use it. If a few large sessions produce most of your returns, a profit target truncates exactly the days that pay for everything else.
A daily loss limit is the least glamorous risk control and one of the few that reliably works, because the failure it prevents is human. It removes the decision to keep going from someone who is currently losing.
A stop is an instruction to send an order, not a promise about what you will get. Most of the time it works approximately as intended. The exceptions are properties of market structure, and no setting changes them.
These two order types are one word apart and produce opposite failure modes. One fails by price, the other by leaving you in the position — and for an unattended system, the second is the one that breaks things.
Break-even stops feel like free protection. They are not. The entry price is exactly where the market is most likely to return before continuing, which means the mechanism exits good trades at the worst point in their path.
A stop left resting after the target fills does not know the position is gone. A person watching a screen would probably notice. A program will not, unless someone wrote the code to look — which is why the automation case for brackets is stronger.
An unlinked stop left resting after a target fills does not know the position is gone. When it triggers, it opens a new one. OCO exists to prevent exactly that, and understanding where the linkage stops helping is the useful part.
The trigger threshold is the least discussed parameter in trailing stop configuration and one of the most consequential. It does not reduce risk — it decides when a position stops being tested and starts being protected.
Single-tier trailing forces one decision to serve a position's whole life. Tiering lets that decision change as the position develops — which is genuinely useful, and is also exactly what makes it easy to fit noise instead of structure.
Most trailing stop guidance is written for equities and applied to options unchanged. The three places it breaks are specific: gamma, theta, and a bid-ask spread that can widen tenfold between your entry and your exit.
Most traders carry a mental model of a trailing stop as a floor. It is not — it is a trigger that sends an order, and the price you get is whatever the market offers. Understanding the difference is what separates a stop that helps from one that surprises you.
Win rate is the metric the trading-education industry loves to advertise, because a high percentage sounds like skill. It is also nearly useless on its own: a 90% win rate can lose money and a 40% win rate can be highly profitable, because what determines profitability is expectancy, the size of wins and losses, not how often you win. Here is the math, and why optimizing for win rate pushes you toward exactly the wrong strategies.
Everyone says to forward-test a strategy to catch overfitting. Almost no one explains how to structure that testing rigorously. Out-of-sample validation and walk-forward analysis are the methods: optimize on data the strategy is allowed to see, evaluate only on data it is not. This explains how they work, the anchored-versus-rolling choice, the data-leakage traps, and the honest limit that even these methods can be gamed.
A backtest result is only as honest as the assumptions behind it, and several common ones systematically make a strategy look better than it is. Tick versus bar data, slippage assumptions, survivorship bias, look-ahead bias, and overfitting each inflate results in a specific way. This is a practical guide to reading a backtest report without letting it fool you.
Backtesting and paper trading are both ways to test a strategy without risking money, and they are not two grades of the same thing. They answer categorically different questions, one about the past you can see, one about live conditions you have not, and each has its own failure mode. Treating them as interchangeable, or treating either as proof a strategy will profit, is how traders talk themselves into confidence they have not earned.
Market, limit, stop, and stop-limit are the core order types, and choosing among them is a tradeoff between certainty of fill and certainty of price. Automation changes the calculus, because software cannot watch a resting order and improvise the way a human can. This explains each order type honestly, including the ways stops do not work the way people assume, and which fit automated execution.
