Market Profile (TPO)

A charting method that organizes price by how much time the market spent at each level, building the familiar bell-shaped profile.

What it is

Market Profile is a charting technique developed by J. Peter Steidlmayer at the Chicago Board of Trade in the 1980s. Instead of plotting price against time left-to-right, it stacks time against price: each half-hour period of the session is assigned a letter (a Time Price Opportunity, or TPO), and every price the market touched during that period gets that letter printed next to it. Stacking the letters produces a horizontal distribution showing where the market spent the most and least time. This is the key difference from Volume Profile, which people often confuse it with: Market Profile measures time at price (how long the market traded at each level), while Volume Profile measures volume at price (how many contracts or shares actually changed hands there). The two often look similar but can diverge meaningfully — a level can be touched for many periods on thin volume, or absorb heavy volume in a single brief visit. Market Profile is a way of organizing what already happened in a session; it is descriptive, not predictive, and is presented here for education only. No charting method forecasts price, and all trading carries risk of loss.

How it works

A session is divided into fixed periods, traditionally 30 minutes each, labeled A, B, C and so on. For each period, one letter is printed at every price increment the market traded through. As the session progresses, letters accumulate: prices visited in many periods build long horizontal rows, and prices visited briefly show only one or two letters. From this distribution several standard references are derived. The Point of Control (POC) is the price with the most TPOs — where the market spent the most time. The Value Area is the range around the POC containing roughly 70% of all TPOs (an analogy to one standard deviation of a normal distribution). Single prints, where only one letter appears, mark levels the market moved through quickly. The Initial Balance is the range of the first two periods. One honesty note that matters for platform users: most charting platforms, including TradingView, do not carry true TPO period data for every instrument — profile tools there frequently reconstruct an approximation from bar highs and lows at some chart resolution, or substitute a volume-based profile entirely. An approximation built from bar ranges assumes the market traded every price within each bar's range for the whole bar, which is not literally true, so reconstructed profiles are estimates, not exchange-grade TPO records.

How traders read it

• Shape: a symmetric, bell-shaped profile is conventionally read as a balanced session where buyers and sellers agreed on value; an elongated, thin profile is read as an imbalanced, directional session. These are descriptions of the completed session, not forecasts.
• Point of Control (POC): the price row with the most TPOs marks where the market spent the most time — often watched as a reference level in later sessions, though nothing obliges price to respect it.
• Value Area (VAH/VAL): the band holding about 70% of TPOs. Some traders note whether price opens or trades inside versus outside the prior day's value area as context for balance versus imbalance.
• Single prints and tails: rows with only one letter mark fast, one-sided movement; long tails at a session extreme are read as rejection of those prices during that session.
• TPO versus volume: compare the time-based profile with a volume profile when both are available — a time-heavy but volume-light level tells a different story than a level with genuine transaction volume.
• Remember the profile summarizes a past session. It assigns no probability to future behavior, and any reading should sit alongside a defined risk plan.

Common settings

The traditional configuration is 30-minute TPO periods over one full exchange session, with a 70% Value Area. Intraday traders sometimes use shorter periods (such as 15 minutes) for finer profiles, or merge multiple days into composite profiles for longer-term structure. The price row size (tick increment per row) materially changes the profile's shape — coarse rows smooth detail away, fine rows fragment it — so row sizing should be stated explicitly when comparing profiles. On platforms without true TPO data, check whether the tool is building the profile from lower-timeframe bar ranges and at what resolution, because that choice changes every level the profile reports.

Strengths

• Organizes a session by where the market spent time, which a standard time-axis chart makes hard to see at a glance.
• Produces concrete, reproducible reference levels (POC, Value Area, Initial Balance, single prints) rather than subjective drawings.
• The time-at-price versus volume-at-price distinction adds context a volume profile alone cannot provide.
• Grounded in a documented methodology with decades of literature, originating from exchange practice rather than marketing.
• Composite profiles across multiple sessions give a structured way to describe longer-term ranges.

Pitfalls to watch

• It is routinely confused with Volume Profile: TPO counts time at price, Volume Profile counts volume at price. Treating one as the other leads to wrong conclusions about where transactions actually occurred.
• Platform approximation: without genuine period-by-period TPO data, platforms reconstruct profiles from bar highs and lows, which assumes price traded uniformly through each bar's range. The resulting POC and Value Area can differ from exchange-grade profiles.
• The developing profile changes all session: POC and Value Area migrate as new TPOs print, so any intraday reading is provisional until the session closes — a form of repaint risk if you backtest on finished profiles.
• Settings sensitivity: period length, row size, and session boundaries (including timezone and DST handling) all reshape the profile; two traders with different settings see different 'key levels'.
• The bell-curve analogy is loose: real sessions are not normal distributions, and the 70% Value Area convention is a heuristic, not statistics.
• No predictive power: a profile describes where trading happened. It does not forecast whether value will be accepted or rejected tomorrow, and no method removes the risk of loss.

Pine v6 example

//@version=6
indicator("Time-at-Price POC (TPO approximation)", overlay = true)

// Approximates a time-at-price profile from bar ranges over a lookback
// window. Honest caveat: this assumes price traded through each bar's
// full range, so it is an estimate, not exchange-grade TPO data.
lookback = input.int(120, "Lookback Bars", minval = 10)
rows     = input.int(24, "Price Rows", minval = 5, maxval = 50)

hi   = ta.highest(high, lookback)
lo   = ta.lowest(low, lookback)
step = (hi - lo) / rows

float pocPrice = na
int bestCount = 0
if step > 0
    for i = 0 to rows - 1
        rowLo = lo + step * i
        rowHi = rowLo + step
        int cnt = 0
        for j = 0 to lookback - 1
            if high[j] >= rowLo and low[j] <= rowHi
                cnt += 1
        if cnt > bestCount
            bestCount := cnt
            pocPrice := rowLo + step / 2

plot(pocPrice, "Time-at-price POC", color = color.orange, linewidth = 2)

Pro tip: Before leaning on any profile level, find out exactly what your platform is counting: true 30-minute TPO letters, a bar-range reconstruction, or volume at price. A POC computed three different ways gives three different prices, and treating an approximation as exchange data is how phantom 'key levels' get traded. Note also that the developing profile migrates all session, so intraday POC reads are provisional. This is educational material only, not financial advice — profiles describe past trading, and all trading carries risk of loss.

Educational only — not financial advice, not a recommendation to trade. No indicator is predictive; trading involves substantial risk of loss.