A map of what people actually do, what each approach assumes, and where the evidence is solid, contested or absent.

This page describes approaches. It does not recommend one, because this site does not do that, and because the honest summary for several of them is that the evidence does not support the confidence with which they are sold.

Slide the picture sideways →

First, the timeframe

Before method comes horizon, and it decides how the costs from the next page fall on you.

Scalping — minutes or seconds, many trades a day. Maximum exposure to fees and spread, and direct competition with automated systems that are faster than any human by orders of magnitude.

Day trading — positions opened and closed within the day. No overnight risk, and no ability to sit out a move you were early on.

Swing trading — days to weeks. Fewer trades, less cost drag, and exposure to whatever happens while you are asleep.

Position trading — weeks to months. At this end the distinction from investing blurs, and the costs stop dominating.

The shorter the timeframe, the larger the share of your result that is decided by costs rather than by judgement. This is not an opinion; it is the arithmetic of paying a fee more often.

Technical analysis

Studying price and volume history to infer what comes next: trends, support and resistance, indicators derived from past prices, named formations.

What is real: levels where a great deal has traded are levels where many people have a pending decision, and that is genuine information about where activity may cluster. Indicators are also a compact way to describe what has already happened.

What is contested: the predictive claim. Studies of chart patterns find weak and inconsistent results, and any effect that does exist tends not to survive trading costs. There is also a self-referential element — a level matters partly because enough people are watching it — which makes the whole thing harder to test than it looks.

The reasonable position is that this is a framework for organising observation, not a forecasting instrument. Anybody presenting it as the latter, particularly while selling access to it, is going well past the evidence.

Fundamental and on-chain analysis

Assessing the thing itself: what the network does, who uses it, how supply is issued, who holds it, what the revenue is if there is any. On-chain data makes some of this unusually checkable — active addresses, fees paid, supply distribution, flows to and from exchanges are all public.

The limitation is timing. Something can be genuinely overvalued for years. Fundamental work tells you about the object; it says very little about the next month, which is what a trader needs.

Arbitrage

The same asset trading at different prices in two places. Buy where it is cheap, sell where it is dear, pocket the difference — in principle without taking a directional view at all.

This is a real, structural edge and it is almost entirely occupied by automated systems that act in milliseconds and hold inventory on both venues. The gaps that survive long enough for a person to act on them usually exist for a reason: a withdrawal is suspended, a transfer takes twenty minutes, or one of the venues is not going to let you take the money out. The apparent free money is a payment for a risk that is not visible on the two price screens.

Market making

Quoting both a buy and a sell price and earning the spread from the flow between them. Also structural, also real, and it is a business — it needs capital on both sides, continuous operation, and management of the risk that the market moves while you are holding inventory. Venues actively court these participants, which tells you it is not an opportunity lying around unclaimed.

Dollar-cost averaging

Buying a fixed amount at fixed intervals regardless of price. Strictly it is not trading — there is no view and no exit plan — but it belongs on the map because it is what many people do instead.

It removes timing from the decision, which removes both the risk of buying everything at a peak and the possibility of doing better than average. It is mechanical, it is cheap if the intervals are not too frequent, and its honest description is that it converts an unanswerable question into a schedule.

Bots, signals and copy trading

Automation is not a method, it is a way of executing one. A bot running a rule executes that rule faster and without hesitation, which is useful if the rule has an edge and merely quicker if it does not.

Published track records deserve the same scrutiny. Displayed performance is usually the surviving accounts, since the ones that blew up stop being displayed. A leaderboard is a filter, and what it filters out is the information you need.