Stop-loss discipline vs. holding conviction: how ACCE rebalance decisions on SRPT, COLO-B, and SNAP reveal when to cut and when to hold.
Stop-Loss vs Conviction: When to Cut a Position
Every investor eventually faces the same uncomfortable moment: a position is down, the thesis feels shakier than it did at entry, and the question becomes whether to hold or walk away. Stop-loss discipline and conviction-based holding are not opposites. They are two tools that serve different situations, and confusing them is one of the most reliable ways to destroy a portfolio.
The ACCE rebalance log offers three recent case studies that illustrate exactly where the line sits.
The Mechanical Stop-Loss: What It Gets Right and Wrong
A hard stop-loss, say 8% or 15% below entry, removes emotion from the decision. That is its entire value. It does not require you to re-analyze the thesis, re-read the filings, or talk yourself into staying. The position hits a level, and you exit.
The problem is that mechanical stops treat all drawdowns as equivalent. A stock down 12% because the sector rotated is not the same as a stock down 12% because the core thesis broke. Treating them identically means you will exit good positions at bad times and miss recoveries that were always coming.
The better framework asks one question before anything else: did the reason you bought the stock change?
If the answer is no, a drawdown is noise. If the answer is yes, no stop-loss is needed because the decision to exit is already made on fundamental grounds.
Case Study 1: COLO-B, a Loss Accepted Early
Coloplast was picked as a healthcare compounder with a switching-cost moat and a margin of safety built into the entry price. The position closed at a loss of 3.0%.
Looking at the current data, COLO-B trades at $402.80 with a trailing P/E of 44.1 and year-over-year revenue growth of just 2.2%. The analyst consensus target sits at $460.90, which implies meaningful upside from current levels. So why close?
The answer is that a 3% loss on a position where the thesis was intact is not a stop-loss exit. It is a reallocation decision. The opportunity cost of holding a slow-growth healthcare name while higher-conviction ideas sit in the queue is a real cost, even if it does not show up on a brokerage statement. Cutting a small loss to redeploy into a better-ranked idea is not capitulation. It is portfolio management.
The lesson: small losses taken early are a feature, not a failure, provided the capital moves somewhere with a stronger setup.
Case Study 2: SNAP, Held Through Volatility to a 40.3% Win
Snap was not an obvious hold. The stock carries no earnings, a forward P/E of 7.6, and an ACCE score of 44/100 as of May 20, 2026. The current price sits at $5.62 against an analyst target of $7.63.
At the time of the pick, the thesis centered on AR-driven revenue diversification and a turnaround in monetization. That thesis did not change during the holding period, even when the stock moved against the position. The result was a 40.3% gain.
This is the case for conviction-based holding. The stock was cheap relative to its monetization trajectory. The business was executing. The score was low partly because trailing metrics lagged the forward story, which is exactly the setup where mechanical stops destroy value. An 8% stop would have ended this trade before it moved.
The lesson: when the thesis is intact and the business is executing, a low score or a rough patch in price action is not a sell signal.
Case Study 3: SRPT, the Open Position That Demands a Framework
Sarepta Therapeutics is the hardest current case. The stock trades at $17.15 with a market cap of $1.81B. Revenue is down 1.9% year-over-year. The ACCE score sits at 37/100. The analyst target is $22.00, which represents meaningful upside, but the forward P/E of 6.3 against a trailing P/E of 49.0 tells you the market is pricing in a dramatic earnings inflection that has not yet arrived.
The pick thesis was built around a gene therapy catalyst and an anticipated FDA outcome. That is a binary event thesis, which operates under completely different rules than a compounder or a turnaround.
Binary event positions require pre-defined exit rules set before the event, not after. If the FDA decision comes in negative or delayed, the thesis is broken regardless of where the stock trades. Holding through a broken binary thesis because you are down is the definition of hope-based investing.
The framework here: define the catalyst, define the outcome that validates the thesis, and define the outcome that breaks it. Then execute mechanically when the event resolves. The current open status means the clock is running.
The Decision Tree in Practice
When a position moves against you, work through this sequence:
- Is the thesis intact? If yes, the drawdown is a price event, not a fundamental event. Hold or add.
- Did the business change? Earnings miss, guidance cut, management departure, competitive disruption. If yes, exit regardless of where the stock trades relative to your cost basis.
- Is this a binary event position? Define the pass/fail criteria before the event. Do not improvise after.
- What is the opportunity cost? A flat or slightly negative position tying up capital that could go into a higher-conviction idea is a real drag. Small losses to fund better ideas are rational.
Cost basis is irrelevant to every one of these questions. The stock does not know what you paid. The only question is whether the forward expected value of holding exceeds the forward expected value of the next best use of that capital.
What the ACCE Log Shows Consistently
Across the rebalance history, the losses that compound into large losses share one characteristic: the exit was delayed because the investor was waiting to get back to breakeven. COLO-B at minus 3% is a rounding error. The same position held another six months on a broken thesis becomes a different conversation entirely.
The wins, including SNAP at 40.3% and IMCD at 40.0%, came from holding through noise when the underlying business kept delivering. Neither of those positions would have survived a mechanical 10% stop.
Discipline is not the same as rigidity. The best exit decisions combine a pre-defined thesis with the willingness to act when that thesis breaks, and the patience to do nothing when it does not. The market will keep testing both.
