[Investment Masters] Stock #653 That Broke Support and Bounced Back—The Reversal Was Already Complete by Close

There's a scene you see all the time in stock communities. A beaten-down stock breaks below its support line, then climbs back above it before the close. And the comments start rolling in. Big players are shaking out retail, this is the time to buy.

This idea traces back a hundred years. Richard Wyckoff called it a spring, and it's still the go-to explanation whenever people talk about accumulation by insiders.

I wanted to know if this was actually true. So I went through twelve years of daily price data for every Korean-listed stock and found every instance that fit the pattern. I ended up with 653 occurrences.

Here's the bottom line: Wyckoff saw something real. But the cost of that move was already being paid at a time we can't touch.

What exactly is a spring?

It's hard to visualize just from the description. A picture will make it click instantly.

After a long decline, there's a sideways phase, and the arrow marks where the price briefly dips below the base before reverting—that's the spring. This is a concept diagram, not actual data.

Wyckoff's logic goes like this: While big money is accumulating stock, the price gets stuck in a narrow band, moving sideways and boring. At the end, they deliberately dip just below support to shake out anyone with stop-losses. Then they reverse course immediately.

So if support breaks but the closing price is back above it by day's end, that's not a collapse—it's a feint. According to this theory, that's when you should buy.

I isolated only the setups he described

Here's what matters: you can't call every support break in every stock a spring. Wyckoff's spring only works within an accumulation phase. If you grab loose criteria and blast through all stocks, that's not verification—that's wasting time.

So I only included candidates that met all three conditions: a significant prior decline, roughly three months of sideways consolidation after that, and an actual intraday break below the base. I excluded any stock that couldn't be traded freely.

Over twelve years, I found about 60,000 days of consolidation that matched these criteria. But the base was actually breached intraday on only 6% of them. Of those real breaks that reversed back above support by close—the true springs—there were 653. The rest broke and stayed broken, hitting genuine breakdowns, and there were over 3,000 of those.

I only used information available by market close that day. No hindsight like "it turned out to be the bottom." I bought at that day's close and measured the result one month later, the way a real trader would.

I figured springs would outperform. The data said the opposite.

The supposed "trick" actually underperformed

When you bought the spring at the close and held for a month, the average return was minus 0.3%. Less than half of them went up. The genuine breakdowns—the ones you're "supposed" to avoid—averaged positive over the same period, and more than half went up.

If you bought at that day's close	Average one month later	% that went up
Spring (dipped then recovered)	-0.30%	48%
Genuine breakdown (close below bottom)	+0.58%	53%

Year by year, it was the same story. Springs outperformed genuine breakdowns in only three out of ten years. I thought maybe I'd set my criteria wrong, so I ran it eight different ways, adjusting box length and width. Springs came out ahead in exactly one combination. That's not data manipulation—that's the result.

So is a hundred-year-old theory just plain wrong? I didn't stop there. I ran one more test, and that's where I found the answer to this whole piece.

The reversal did happen. It just ended by close

I recalculated using the intraday low instead of the closing price. You obviously can't buy at that exact price, but it tells you whether a reversal actually occurred.

The red bars are measured from that day's low; the gray bars are measured from the closing price we can actually buy.

Measuring from the low, springs came in at +3.1%, nearly double the genuine breakdowns. This pattern repeated identically in nine out of ten years. That's not coincidence.

The red bar is higher than gray in every year except 2025. Even in 2020 when returns were negative, springs lost less.

So Wyckoff's observation was correct. There was definitely reversal power in that spot. The question was when that power ran out.

On the day a spring occurred, the price had already recovered an average of 3.5% from that day's low to the close. When a genuine breakdown happened, the intraday recovery was only around 1%. For something to qualify as a spring, the reversal has to complete by close, so that makes sense.

This is the crux of it. The moment we look at the chart after hours and say "oh, that's a spring," the rally is already priced in. We're paying the full cost of that intraday move just to confirm the signal. That's why we end up with losses.

Why did the other side perform marginally better? There's one more thing to explain here.

The genuine breakdowns' slight edge follows the same logic. On those days, the bounce hasn't happened yet—it comes over the following days. But that's my interpretation. The only facts I have are the returns for both groups. And the difference is less than 1% over a month, so don't take this as a signal to buy breakdowns.

So what do you actually do?

The spring wasn't a false theory—it was a late signal. That distinction matters. You discard a false theory, but a late signal just needs a different approach.

First, this technique didn't work when you confirmed it at the close and entered. For an office worker who can't watch intraday and sees a community post later, twelve years of data showed no edge.

Second, the real value happened at that exact moment the support line broke. To profit, you'd need to place an order at that price ahead of time, not wait for the reversal. But remember, true breakdowns outnumbered springs massively. 653 to over 3,000.

Third, this isn't unique to springs. Any signal visible in a single candle is already priced in by the time you see it. I ran into the same wall with bullish alignments and new highs. I wrote an article analyzing bullish alignments over twelve years, and it hit the same ceiling.

In one sentence: I don't know if big players shook out retail. But the data is clear on this: by the time you recognize that setup, you've already paid the price.

This article is a record of historical price data and is not a recommendation to buy or sell any specific stock. Investment decisions and their outcomes are your own responsibility.