When to Buy and Sell Stocks: What the Evidence Actually Supports

The premise, corrected

“When is the best time to buy?” assumes timing is the variable that determines outcomes. For almost every participant it is not. Position sizing, transaction costs, and whether you exit according to a plan dominate results so heavily that timing differences are usually lost in the noise.

That said, the trading day and the calendar do have structural features worth knowing — some genuinely mechanical, others folklore that has survived through repetition. Separating them is useful.

Structural features of the trading session

These are tendencies, not rules, and they vary by instrument and by day. The genuinely actionable point is narrower than it looks: spreads and slippage are worst at the open, which is a cost fact rather than a directional one, and it is enough to justify handling the opening period deliberately.

Calendar effects: what survives scrutiny

Calendar anomalies — day-of-week effects, month-turn effects, January effects — are extensively documented in academic literature. Three cautions before treating any of them as tradeable:

• Erosion. Anomalies that become widely known tend to weaken. An edge that is public and easily traded is competed away, which is the ordinary functioning of a market.
• Costs. Most documented calendar effects are small relative to real transaction costs. A 0.15% average edge does not survive brokerage, taxes and slippage.
• Multiple-testing. With enough calendar slices — days, weeks, months, pre-holiday, post-holiday — some will appear significant by chance alone. Many published anomalies are artefacts of searching.

The honest position: interesting, largely not actionable, and any pattern you intend to use should be tested on your own data with your own costs before it becomes a rule.

Expiry: the one genuinely mechanical effect

Unlike calendar folklore, derivatives expiry has real mechanical consequences that do not erode, because they follow from how the instruments are constructed:

• Time decay accelerates as expiry approaches, which changes the economics of every option position regardless of direction.
• Gamma rises sharply for near-the-money strikes, so small index moves produce disproportionate option price changes. Positions become far more sensitive in their final hours.
• Rollover activity distorts open-interest readings during expiry week, so positioning data needs interpreting differently.
• Expiry-day behaviour in near-the-money short-dated options resembles a coin flip with costs attached more than it resembles a strategy.

Expiry cycles and contract specifications are set by the exchange and have been revised on several occasions. Confirm current specifications from the NSE contract specification page rather than relying on any provider's description.

For investors: time in, not timing

For long-horizon investors, the weight of evidence points consistently in one direction: attempting to time entries and exits has historically cost more than the declines it avoided. The mechanism is well understood — exits are usually mistimed, and re-entry is delayed by the same caution that prompted the exit, so the recovery is missed.

• Systematic, regular investment removes the timing decision entirely, which is its main advantage.
• Rebalancing on a schedule is a rules-based way of buying weakness and selling strength without forecasting.
• The decision that matters is asset allocation and holding period, not entry date.

This applies to investing. Trading is a different activity in which timing is intrinsic — but even there, the timing that matters is defined by a rule, not by a judgement formed in the moment.

The harder question: when to sell

Far more damage is done by exit decisions than entry decisions, and almost all of it comes from deciding the exit while holding the position.

• Sell at your invalidation — the level or condition at which the premise is disproved, defined before entry and not moved afterwards.
• Sell at your target, or according to a trailing rule that was decided in advance.
• Sell on time if neither occurs. A position that has done nothing for its expected holding period is capital with no thesis attached.
• Do not sell because the position is uncomfortable at a level that does not invalidate anything. This single behaviour — exiting winners early while holding losers — inverts the win/loss ratio most strategies depend on.

Building rules instead of guessing timing

1. 01Define the condition, not the clock“When my setup occurs” is testable. “Around 10am” is a habit.
2. 02Measure execution cost by time of dayIf your strategy triggers at the open, slippage is part of its edge calculation, not an afterthought.
3. 03Write the exit before the entryTarget, invalidation and time-based exit, all three, before the position exists.
4. 04Handle expiry explicitlyEither exclude expiry-day trading or build rules specifically for it. Do not let it happen to you.
5. 05Test any timing rule on your own dataWith your own costs. Most published timing edges do not survive that step.

EqtPulse's strategies specify entry conditions, invalidation and holding period as rules rather than judgements, and publish average holding time, trade frequency and drawdown for each — so the timing question is answered by the framework rather than in the moment.

Frequently asked questions

Related reading