Short-term trading sits on three layers. The instruments: exchange-traded funds and what each one actually holds. The tools: indicators that summarize past prices, which you need to understand as calculations rather than forecasts. And the arithmetic: costs, correlations, leverage and position size, which decide results more often than any signal does. This guide follows that order and ends with the published record of active traders and the scams aimed at them. Orientation for study, not investment advice, and not affiliated with any fund sponsor, index provider, exchange or regulator. Figures were computed from daily prices or read from the cited sources as of November 2026; check the current rule before acting on a number.
Each chapter opens with the short version. Tap one to read the detail.
What an exchange-traded fund is, and what the big ones hold
~2 min
An exchange-traded fund trades all day at a market price, kept near the value of its holdings by large firms that can create and redeem shares. The ticker tells you nothing; the index it tracks tells you what you own.
Such a fund pools money like a mutual fund, but you buy and sell its shares from other traders on an exchange. Two prices matter. Net asset value (NAV) is the holdings' value per share, struck at the close. The market price is whatever trades. Authorized participants, usually large broker-dealers, keep the two close: when shares trade above the basket's value they create new shares and sell them, and when shares trade below it they buy and redeem them. That anchor is strongest when the holdings trade easily.
Then read the index. SPY, IVV and VOO all track the S&P 500 of large US companies. QQQ tracks the Nasdaq-100, which leans heavily on technology. IWM tracks the Russell 2000 of small companies, EFA developed markets outside the US and Canada, and sector funds such as XLK and XLE slice the large-company index by industry.
The trap is counting tickers as diversification. From 2015 to 2019, SPY's daily returns correlated 0.93 with QQQ's, 0.87 with IWM's and 0.83 with EFA's. Four funds, but on most days one direction. And because these indexes weight companies by market value, a few giants carry much of each fund.
Rule: before trading a fund, name its index and its weighting, and check how it correlates with what you already hold.
Funds that do not behave like their label
~2 min
Leveraged and inverse exchange-traded funds promise a multiple of one day's return, and compounding breaks that multiple over longer holds. Bond funds fall when rates rise, and long-maturity funds fall much further than short ones.
A 2x fund seeks twice an index's return for a single day, then resets. Hold it for two days and you meet compounding. If the index gains 10% and then loses 10%, it ends down 1%; the 2x fund gains 20% and loses 20%, ending down 4%. In a market that chops back and forth and goes nowhere, a leveraged fund bleeds: ten alternating days that leave an index flat leave a 2x fund down about 2.4%. Inverse funds suffer the same way. Over those two days a −1x fund also ends down 1%, so if you bet on the fall, you lost even though the index fell.
Bond funds carry a different surprise. When market rates rise, prices of existing fixed-rate bonds fall, and long-maturity bonds fall more. From August 2020 to October 2023, the long-Treasury fund TLT lost 48.4% on dividend-adjusted prices while the one-to-three-year fund SHY lost about 3%. Both hold US government debt, so the credit was never the problem; the rate risk was. A bond fund also never matures, so waiting does not guarantee a return to par.
Commodity products, including gold trusts, sit outside the law that governs ordinary funds and lack some of its protections.
Rule: read a product's objective and holding period before its ticker, and treat "daily," "inverse" and "long-dated" as risk words.
Volatility, the VIX and price bands
~2 min
Volatility is the spread of returns, and it changes a lot from year to year. The VIX is a 30-day volatility expectation implied by option prices, and products tied to it hold futures that tend to lose value over time.
Volatility is usually the standard deviation of daily returns, scaled to a year. It is a property of a period, not of a fund: SPY's figure was 6.7% in 2017 and 33.3% in 2020. Realized volatility measures what happened. Implied volatility is backed out of option prices, so it reflects what traders are paying for protection.
The VIX packages that idea for the S&P 500: the 30-day volatility implied by a range of index options. It moves opposite to stocks on most days; from 2010 to 2024, SPY's daily returns and the VIX's daily changes correlated −0.80. That makes it tempting as a hedge.
Here is the trap. You cannot buy the VIX itself. Volatility funds hold VIX futures, and research on the futures curve finds it usually in contango, with later months priced above nearer ones. To stay near a one-month maturity, a fund keeps selling the cheaper expiring contract and buying the dearer next one, losing a little each roll. Between spikes it bleeds, so a fund that jumps in a crash can still lose money over a year.
Bollinger Bands apply the same statistics to a chart: a moving average with lines, commonly two standard deviations, above and below. A close outside a band says the move was large for recent conditions, nothing more.
Rule: quote volatility with its period, and before holding a volatility product, check what it owns and what the roll costs.
Reading a price chart without fooling yourself
~2 min
Each bar or candle records an open, high, low and close for one period. Patterns and support lines are easy to see in hindsight, so a chart claim only counts once it is an exact, tested rule.
The body of a candlestick runs from the open to the close, and its shadows reach the period's high and low. The interval changes everything: a spike on a five-minute chart vanishes on a weekly one, and an indicator set to 14 periods means minutes on one chart and weeks on another. On a log scale, a rise from 10 to 20 looks the same as one from 50 to 100, since both are +100%; a linear scale makes later moves look larger. For returns, use adjusted prices, which fold in splits and dividends. A 2-for-1 split halves the raw price without anyone losing a cent.
Gaps deserve attention. When news lands outside trading hours, the next open can sit far from the last close with no trades between, and an order waiting inside that gap cannot fill at its price.
The trap is reading meaning into shapes. Support and resistance are zones where price stalled before; whether it stalls there again is an empirical question. Two analysts can draw different trendlines on the same chart, and a head-and-shoulders pattern is often recognized only once it is complete. Named candlestick patterns number in the dozens, and naming one is not evidence that it predicts anything.
Rule: state the interval and scale of any chart you read, and turn any pattern you believe in into a written rule before trusting it.
Moving averages, MACD and trend strength
~2 min
Moving averages smooth past closes and always lag; crossovers mark when two averages change order. Moving average convergence divergence (MACD) and the average directional index are built from the same smoothing and measure the gap between trends and a trend's strength.
A simple moving average is the plain mean of the last N closes: for 10, 11, 12, 11 and 13, the five-day average is 11.4. An exponential moving average (EMA) closes a fixed share of the gap to each new close, and that share is 2 ÷ (N + 1). With N = 5 and a new close of 14, it moves from 11.4 to about 12.27. Every average lags. A simple average's data are on average (N − 1) ÷ 2 periods old, so a 20-day line is reporting from about 9.5 days ago.
A golden cross is a short average crossing above a long one, and a death cross the reverse. MACD subtracts a 26-day EMA from a 12-day EMA and smooths that difference again into a signal line. True range takes the largest of the day's range and the gaps from yesterday's close, and average true range (ATR) smooths it into a typical move size. The average directional index rises in any strong trend, up or down.
The trap is the names. "Golden" makes you hear a forecast, but the event is two averages of the same past prices swapping places. In a sideways market, crossover rules flip back and forth and pay costs on every flip. In a study of Southeast Asian indexes, even the profitable MACD rule won fewer than half its trades; it made money because average wins beat average losses.
Rule: describe a trend indicator by its inputs and its lag, and judge a trend rule by its payoff ratio after costs, not by its hit rate.
Oscillators and volume tools
~2 min
The relative strength index (RSI) and the stochastic oscillator measure how far recent moves have run, on a 0 to 100 scale. On-balance volume and the volume-weighted average price summarize trading volume. None of them says what happens next.
RSI weighs average gains against average losses across a window, usually 14 periods: 100 × average gain ÷ (average gain + average loss). For five daily changes of +1, −0.5, +1.5, +1 and −0.5, the average gain is 0.7 and the average loss 0.2, so RSI is about 77.8. The stochastic %K places today's close within the recent range: a close of 48 in a range from 40 to 50 reads 80. Both rest on a contrarian idea, that sharp moves tend to reverse, which is the opposite assumption from trend tools.
Volume tools work differently. On-balance volume adds the day's volume on an up close and subtracts it on a down close, ignoring the size of the move. The volume-weighted average price is total traded value divided by total shares: 100 shares at $10, 300 at $10.50 and 100 at $11 give $10.50. It is a fair yardstick for judging a fill, and it says nothing about whether the trade was wise.
The trap is the word "overbought." An RSI above 70 says recent gains outweighed losses; in a strong uptrend it can stay there for weeks, and if you sell at every 70 you keep stepping in front of the trend. Platforms also average gains differently, so the same chart can show different RSI values.
Rule: read an oscillator as a description of the last window, match its settings before comparing readings, and test any threshold rule rather than trusting its label.
What the evidence says about trading rules
~2 min
Research on trading rules is mixed: gross profits often appear in tests, mostly vanish after costs, and shrink in newer and more liquid markets. A backtest is believable only after costs, out of sample, and with every tried variant counted.
If prices already reflect their own history, rules built only on that history cannot earn risk-adjusted profits. That is weak-form efficiency, and testing rules is how researchers check it. A widely cited review of 95 studies published from 1988 to 2004 counted 56 finding profits, 20 finding none and 19 mixed, with profits common until about the early 1990s and later work pointing to decline.
Costs decide most cases. In a study of five Southeast Asian stock indexes from 2000 to 2013, several rules showed statistically significant gross returns, but after round-trip costs of 0.5% to 1.5%, most did not earn net returns. Rules did well in Thailand and earned nothing significant in Singapore, the most mature market tested. The breakeven cost of a rule is the extra return it earns per signal; if that is 0.3% and a round trip costs 0.5%, the rule loses money in practice.
The trap is the backtest itself. Tune a window on past data and you have fitted that data. Try fifty rules and keep the best, and chance alone hands you a winner; bootstrap tests such as White's Reality Check exist to adjust for that search. Look-ahead bias uses information no one had at the time, and survivorship bias quietly drops the companies that failed.
Rule: before believing a trading result, ask how many variants were tried, whether the test data played any part in choosing it, and what it earned after realistic costs.
How assets move together, and why that changes
~2 min
A correlation is a number for one window in one regime. Stocks and bonds, growth stocks and interest rates, and nearly all stocks in a crash have shown relationships that changed sign or strength when conditions changed.
Correlation ranges between −1 and +1 and shows how consistently two series move together, not how far. Always correlate returns, not price levels: SPY and GLD prices correlated +0.53 from 2016 to 2019 because both drifted up, while their daily returns correlated −0.21. And always state the window. Inside 2019 alone, a 20-day SPY and TLT correlation ranged from −0.79 to +0.27.
The stock–bond relationship shows why the period matters. SPY and TLT daily returns correlated −0.46 in 2019 and +0.08 in 2022, when both fell together. A Federal Reserve note traces the flip to inflation: when inflation is anchored, bad news for stocks lets central banks cut rates and bonds rally; when inflation runs hot, bad news can come with tighter policy. Watch the sign convention, since correlating stocks with yields flips the sign of correlating stocks with bond prices.
Other pairs moved too. QQQ's returns correlated +0.31 with 10-year yield changes over 2015 to 2019 and −0.16 in 2022. Energy stocks tracked crude oil at about +0.6 in both 2010 to 2019 and 2021 to 2024. Gold ran about −0.3 against a broad dollar index over 2010 to 2019.
The trap that costs most is the crash. The average correlation among nine sector funds was 0.54 in 2019 and 0.88 from late February to April 2020. Diversification thins out exactly when you want it.
Rule: quote every correlation as "in this period, under these conditions," and size positions as if correlations rise toward 1 in a sell-off.
What happens to an order in a fast market
~2 min
The price on your screen is not your fill. A stop order becomes a market order when triggered, so gaps and thin markets can carry its fill far past the stop, while a stop-limit may not fill at all.
A market order takes the best prices available when it arrives. If only 300 shares sit at $12.00, an order for 800 takes them and then the next offers up, so part fills higher. The difference between the price you expected and the price you got is slippage, and it comes from the spread, from movement while the order travels, and from walking through price levels.
Stops are triggers, not guarantees. Once the stop price is reached, a stop order becomes a market order. Suppose you hold 100 shares bought at $50 with a sell stop at $48, planning a $200 loss. If bad news sends the next open to $45, the stop fills near $45 and the loss is about $500. A stop-limit order caps the price but can be skipped entirely: a sell with a $25 stop and a $24 limit will not sell below $24, even if the price falls straight through. Brokers also differ on whether a trade or a quote triggers a stop, and a brief intraday dip can trigger one far below where the stock closes.
Small costs compound. At 0.1% per round trip, 250 round trips a year cost about 25 percentage points of the capital traded. Many strategies that look strong before costs do not survive them.
Rule: plan for the fill, not the quote; treat a stop as a way to exit, not a price promise, and count spread and slippage in every test.
Position size and the arithmetic of losses
~2 min
Size a position from where you will be proven wrong, not from conviction. Expectancy, the average result per trade, matters more than win rate, and deep losses need disproportionately large gains to recover.
Fixed-fractional sizing sets shares so that hitting your stop costs a set slice of the account: shares = account × risk fraction ÷ (entry − stop). With $20,000, 1% risk, an entry at $50 and a stop at $48, that is $200 ÷ $2 = 100 shares. A wider stop means fewer shares. Measuring outcomes in multiples of that initial risk, called R, makes trades comparable.
Expectancy combines how often you win with how much. Winning 40% of trades at 3R and losing 60% at 1R averages +0.6R per trade before costs. The reverse also happens: winning 70% at +1R and losing 30% at −3R averages −0.2R. A high win rate can hide a losing method, which is why the Southeast Asian rule study found its profitable rules winning fewer than half their trades.
Losses compound against you. Recovering from −20% needs +25%, from −50% needs +100%, from −80% needs +400%. SPY's dividend-adjusted price fell 33.7% between February 19 and March 23, 2020; a backtest's worst drawdown is a floor for what to expect, not a ceiling.
The trap is counting positions instead of exposure. Five trades each risking 1% in one sector can lose close to 5% together when correlations jump in a sell-off.
Rule: decide the stop first, size from it, judge a method by expectancy after costs, and add up risk by what moves together.
Margin, short selling and the rules for frequent traders
~2 min
Borrowing magnifies losses on your own money, and a short's loss has no ceiling. Frequent traders also face margin and settlement rules, which changed in 2026, and tax rules that treat their gains as short-term.
Buy $10,000 of stock with $5,000 of your own money and a 20% fall costs $2,000, which is 40% of your equity. Under the Federal Reserve's Regulation T you may borrow up to half the purchase price, and the Financial Industry Regulatory Authority (FINRA) requires at least 25% equity afterward, often more at firms. A margin call can be met by selling your securities without asking you, and you can lose more than you deposited. A short seller borrows shares and sells them; if the price rises from $40 to $50, the loss is $10 a share and has no upper limit, and borrow fees and dividends owed add to it.
The day-trading rules are in transition. The old pattern-day-trader rule applied to four or more day trades in five business days and required $25,000 in equity. FINRA's new intraday margin requirements replaced it from June 4, 2026, with firms allowed until October 20, 2027 to switch, so your firm decides which applies. In a cash account, buying and then selling before paying, or selling a stock bought with unsettled proceeds, triggers violations and restrictions.
Taxes follow holding time. Gains on assets held one year or less are short-term and taxed as ordinary income. Sell at a loss and buy substantially identical shares within 30 days before or after, and the wash-sale rule disallows the loss, adding it to the new shares' cost basis.
Rule: compute the worst case on your own money before borrowing, confirm which margin framework your firm runs, and compare strategies after tax.
The track record, the psychology and the scams
~2 min
Large studies find most individual traders underperform, and day traders mostly lose. Trading shares features with gambling, and scams target the hope of an easy edge with guarantees, bots and hype.
A study of every individual investor on the Colombian Stock Exchange from 2006 to 2016, over 5.3 million trades, found risk-adjusted returns about 4% a year below the market's before costs, and the most active traders did worse even then. Brazilian data show that 97% of people who kept day trading for more than 300 days lost money, and the chance of profit fell the longer they persisted. The Securities and Exchange Commission (SEC) warns that day traders typically suffer severe losses in their first months.
Psychology helps explain why people continue. Researchers describe problematic trading that resembles gambling disorder: small early wins, then chasing losses, poor risk judgment and an illusion of control. In one Australian survey, over 90% of day traders also gambled. The disposition effect, selling winners too soon and holding losers too long, pulls the other way from good risk control, though professionals in one experiment showed that holding losers is rational for assets that revert to a mean.
Scams aim at all of this. A promise of high guaranteed returns with little risk is a classic warning sign, and "AI that cannot lose" is the newest version. Pump-and-dump promoters hype a small stock online, sell into the rise and leave late buyers holding it. Check that any firm or adviser is registered, and verify urgent requests through contacts you look up yourself.
Rule: ask for the base rate before believing a success story, write your exit rules before the trade, and treat any guarantee as a reason to walk away.
Written by Keentune. We are not affiliated with or endorsed by the organizations whose documentation informs this guide, and any linked sources belong to their respective owners.
All exam, test, and product names and trademarks are the property of their respective owners and are used here for identification and reference only. Keentune is independent study practice — not affiliated with, authorized, or endorsed by any of these organizations.