THE PARABOLIC LONG SETUP.

A parabolic long is the mirror image of the parabolic short. A stock falls so far, so fast, that it ends up stretched way below its moving averages, and the sellers who drove it there simply run out.

You buy when it turns, and you sell into the bounce, usually somewhere near the 10-day moving average. It's a quick trade, measured in days. It isn't a bet that the stock has bottomed for good.

What draws people to it is speed. When a capitulation reverses, the snap back can be violent. The catch is that nothing forces it to happen. A crashed stock can keep on crashing, and the stop on this setup is usually wide because the stock is moving so much.

Look at enough real examples and one pattern stands out. The bounces that worked tended to come when the whole market was washing out at the same time.

Why crashed stocks snap back

When a stock falls for days without a pause, it's being sold by people who have to sell. Think margin calls, stop orders, funds cutting risk, and holders who just can't take another red day.

That selling isn't a judgement on the company. And once it's exhausted, there's nobody left to sell at those prices, so the first buyers who show up can move the price a long way (with some help from short sellers taking their profits).

Kristjan Kullamägi (Qullamaggie) treats the parabolic long as the reverse of his parabolic short: the same rubber band, stretched the other way. In his framework, a stock that went parabolic and then collapses 50–60% or more within a few days can bounce hard, sometimes 50–100% within a few days. His published long example is GBTC in 2017.

He's also clear that the two sides aren't equal. Per community notes of his streams, he called parabolic longs "a different game". A parabolic stock that's crashing tends to fall back towards its averages, but a crashed stock is under no obligation to bounce.

Sometimes a company falls 60% because something is genuinely wrong.

What counts as "parabolic down"

"Down a lot" isn't the setup. What separates a capitulation from an ordinary downtrend is the same four traits that define a blow-off top, just turned upside down:

• Size of the fall. Qullamaggie's bar is a collapse of 50–60% or more in days. That's rare in liquid stocks outside of market crashes, so most traders who study the setup look at smaller versions too. A practical scan is 30% or more below the 10-day high, and the deeper the fall, the closer it is to his picture.
• Consecutive down days. You want a string of lower closes, three or more, rather than one gap on bad news. A single earnings gap is a repricing, not a capitulation, because there's no crowd of forced sellers to run out.
• Extension below the moving averages. Measure the gap between the price and the 10-day line in multiples of the stock's average daily range (ADR), so you can compare volatile and quiet stocks fairly. MNDY in March 2022 closed 1.9 ADRs below its 10-day average, and PLNT in March 2020 closed almost 9 below it.
• Panic volume. Volume that spikes as the price collapses, which tells you the forced selling is happening right now. That's the classic description, but treat it as a description rather than a filter. MNDY's last two down days in March 2022 traded at about its normal volume, and it still bounced 70% in six sessions.

Here's one example from the notes of his streams. RMO in January 2021 was down five straight days and about 52% off its high when he bought it on the opening-range high, and he rated it a five-star mean-reversion setup.

The entry: wait for the turn

As with the short, the stretched chart is only the setup. The trade starts when the stock shows you it has stopped going down. Qullamaggie's triggers are intraday:

• Opening-range high. Buy when the price trades above the high of the first 1- or 5-minute candle. It's the same opening-range-high entry he uses for breakouts.
• The first green 5-minute candle. If the stock keeps falling from the open, wait for the first 5-minute candle that closes green, and buy the break of its high.
• A break of a tight intraday range. Per his stream notes, he likes to see the selling turn into structure first (for example, a few green 60-minute candles making higher lows) and then a break out of that range.

The same notes add two refinements. First, speed is the whole edge, so the bounce has to come straight away. If the stock drifts sideways for days, the setup is dead.

Second, a big gap up in your direction removes a lot of the stretch you were trying to buy, so he sees the odds as lower.

You'll notice that buying a stock because it's red for the sixth day running isn't on this list. That's just a guess about where the bottom is.

The stop: the low of the day

Neither of the Qullamaggie sources we use gives a stop that's specific to the parabolic long. What they give is his general rule for his setups: a stop at the low of the day, never wider than the stock's ADR.

On an opening-range entry, that means the low of the day, which is often the low of the opening range. It's what the examples below use, but bear in mind it's a general rule applied to this setup, not a setup-specific one.

The practical problem is width. A stock that has fallen 40% in a week can have an ADR of 10% or more, and the low of day can be further away still.

In three of the six examples below (PLNT, AFRM and AMD), the low of day was more than one ADR from the entry, so the cap kicked in. AFRM's stop still ended up 12% below its entry.

A wide stop isn't a reason to skip the trade, though. It's a reason to size it down.

The exit: sell into the bounce

For the parabolic short, Qullamaggie's targets are the 10- and 20-day moving averages. Our sources don't state targets for the long side, but the mirror image is the natural plan, and it's what the examples below use: sell into the 10-day average, or sell half there and the rest at the 20-day.

At the low, the moving averages are far above the price, which is what makes the trade worth taking. But they also fall every day the stock stays low, so the target comes down to meet the price.

When the profit arrives, take it. A capitulation bounce is a reflex, not a new trend, and the stock often tests the low again within weeks. PLNT (below) gave back a big slice of its bounce within ten days.

It's also worth adding a time stop. If the stock hasn't reached the 10-day within about five sessions, the bounce hasn't come, and by Qullamaggie's own speed rule the setup is over.

Sizing: smaller than your normal trade

Our sources give no sizing rule for this setup, so this part is TradePiko's reasoning rather than anyone's published rule. Three things argue for a smaller position than your usual breakout:

• The stop is wide, so the same dollar risk buys fewer shares. Sizing by the stop takes care of this automatically (see position sizing for swing trading).
• Gaps are common. These stocks move 10% overnight, so the stop can be gapped through, and the loss can end up bigger than the planned 1R.
• The win rate is modest, and the losers come in clusters, because crashes cluster.

Short sellers make the same point: go smaller in higher-ADR names. Half of your normal risk per trade is a reasonable place to start.

How it differs from the parabolic short

Criteria, triggers and targets as described by Qullamaggie; the long-side stop and target are the mirror of his short rules. See the parabolic short post for the short side in full.

In one sense the long is the safer of the two. Your loss can't be more than what you paid, and there's no borrow and no squeeze.

In another sense it's harder. In our study of parabolic runs, 98% touched the 10-day average within 20 sessions. Crashed stocks don't come with anything like that reliability.

Parabolic long mistakes to avoid

• Catching the knife. Buying on the way down because it "has to bounce". Wait for the turn.
• Treating a news gap as capitulation. One gap on bad news isn't a run of forced selling.
• Full size on a wide stop. The stop is usually wide, so size it down.
• Holding for the round trip. The target is the 10-day, maybe the 20-day. It isn't the old high.
• Trading it in a calm market. A deep fall in one stock is a story about that stock. A deep fall everywhere is a story about forced selling, and that's the version that tends to reverse.

A real example: PLNT, March 2020

Planet Fitness (PLNT) is about as clean a capitulation as you'll find. As the market crashed and gyms closed in March 2020, it fell 65% in nine sessions, with three lower closes in a row into the low.

By then it was almost nine ADRs below its 10-day average, volume had hit five times normal, and SPY was 29% off its high. So this was a broad washout, not one stock in trouble.

The next morning it traded above the previous day's high, which is the daily version of the trigger (later and cruder than an opening-range high, but you can check it from daily data). The low of the day was more than one ADR away, so the stop went one ADR below the entry instead.

The sale came at the 10-day average three sessions later, for +4.0R. The bounce kept going to 124% off the low, the far end of the range Qullamaggie describes. Then it gave a lot of it back and closed below the exit within ten days. Taking profits into the average kept the gain, and waiting for more would have given much of it up.

Not just tech: VST, March 2020

The March 2020 crash didn't only hit growth stocks. Vistra (VST), a power company, fell 46% in nine sessions into its low, with three lower closes in a row.

By the setup day it had closed 4.6 ADRs below its 10-day average, and SPY was 29% off its high. Same washout as PLNT, a very different kind of company.

The next morning it traded above the previous day's high. This time the low of the day was inside one ADR, so the stop simply went there.

The stock chopped for two days but held above that low, and it reached its 10-day average three sessions after the entry for +2.5R.

A second leg down: MNDY, March 2022

Monday.com (MNDY) had already been cut down in February 2022, bounced, and then rolled over again in March. It fell 34% in nine sessions while the whole market was in a correction.

What it didn't have was panic volume. Its last two down days traded at about normal volume, a good reminder that the volume spike describes a capitulation but isn't a requirement.

It turned the next day, trading above the prior day's high. The low of the day was inside one ADR this time, so the stop simply went there, and the stock reached its 10-day average one session later for +1.7R.

And this bounce kept going, to 70% off the low in six sessions. You can't know in advance which ones will run like that, which is why the first sale still goes at the 10-day.

Waiting for the turn: AFRM, May 2022

Affirm (AFRM) shows why you wait for the trigger. In May 2022 it had the setup: three lower closes, 37% off its 10-day high, and a whole market washing out around it.

But the turn wasn't there. For two more days it never traded above the prior day's high, and buying the red day would have left you 29% under water.

The trigger finally came on the day of the low, which ended up 58% off the late-April high, the size of fall Qullamaggie describes. The low of the day was about two ADRs away, so the stop went at one ADR.

The next morning AFRM opened above its 10-day average, so the sale came at the open for +3.4R. When a stock gaps past your target, the plan has already done its job (and it didn't go straight up from there, either, falling back later in May before it bounced again).

A failed pop: ROKU, June 2022

Roku (ROKU) is a messier chart, and a useful one. In June 2022 it jumped on heavy volume, then gave all of it back and more. Three straight lower closes left it 31% below its 10-day high, as the whole market sank into its June low with SPY 19% off its high.

Like AFRM, it didn't turn straight away. The day after the setup it made a slightly lower low and never got above the prior day's high.

The trigger came the session after the low. The low of the day was about half an ADR under the entry, so that's where the stop went, and three sessions later the stock reached its 10-day average for +2.1R.

The bounce kept going to 40% off the low before it faded again, another reminder that a capitulation bounce is a reflex, not a new trend.

A washout bounce: AMD, April 2025

The April 2025 tariff sell-off produced the same pattern on a bigger scale. SPY fell 12% in four sessions, and AMD fell 34% in nine, with four straight lower closes.

When the US paused most of its new tariffs and the market turned, AMD traded through the previous day's high. The low of the day was more than one ADR away, so the stop went at one ADR, and the 10-day average came the same afternoon for +1.8R.

Finding your own version

If you trade this setup, or want to, three questions decide whether it belongs in your playbook:

• Where were your entries? Pull up your past bounce trades and check whether you bought a turn (a reclaim of the prior day's high, or an opening-range high) or just bought a red day.
• Did you take the 10-day? Look at the MFE on each trade, which is the most it was ever in profit. If your bounce trades regularly reached the 10-day and you gave the gain back, the problem is the exit, not the setup.
• What was the market doing? Split your results by whether the index was in a sharp correction at the time. If your profitable bounce trades all came in washouts, trade it only then, and use a simple market filter to tell you when that is.

Then study examples. Qullamaggie built his edge by going through thousands of past charts, and capitulations are rare enough that you have to collect them on purpose: March 2020, the 2022 bear market, April 2025.