Scaling In vs All-In: How My SOXL Bot Sized 50 Buys Across 7 Months

One of the biggest decisions in any strategy isn't what to buy — it's how much, and when. Do you commit your full position in one shot, or scale in across multiple purchases? My automated SOXL bot answered this concretely: over seven months, it made 50 separate buy orders, never going all-in at once. Here's the real breakdown of how it sized positions, why tranche-buying makes sense on a 3x leveraged ETF, and the specific month where the approach backfired.

Disclosure: Personal, small-scale experiment for educational purposes only. Not investment advice. Leveraged ETFs carry extreme risk.

The Monthly Buy Cadence

Here's how those 50 buys were distributed across the run:

Month	Buy orders
Dec 2025	2
Jan 2026	6
Feb 2026	10
Mar 2026	7
Apr 2026	3
May 2026	6
Jun 2026	14
Jul 2026	2

The bot bought across a huge price range — from a low of $42.07 to a high of $273.21 — never trying to time a single perfect entry. Instead, it added in tranches whenever its pullback signal triggered, building and rebuilding positions as conditions allowed.

Why Scale In on a 3x ETF

Scaling in exists to solve one problem: you don't know where the bottom is, and on a leveraged ETF, being wrong all at once is devastating.

If the bot had committed its full capital on a single entry and that entry was poorly timed, a 3x ETF could have handed it a 20-30% loss on the entire position within days. By splitting entries into tranches, no single bad entry could sink the whole position. Some buys landed well, some landed poorly, and the blended average price was far more forgiving than any single all-in bet would have been. On an instrument this volatile, that diversification across time is a core risk control, not a minor tweak.

Where It Helped

The early months show scaling-in at its best. Through December, January, and February, the bot accumulated across the $42-$72 range in 18 buys. Because it never committed everything at one price, it built a position with a reasonable blended cost — and when SOXL later ran to the $280s, that patiently-accumulated base produced the strategy's biggest winners (+32%, and later the +47%/+45%/+77% cluster).

Where It Backfired: June's 14 Buys

Now look at June: 14 buy orders in a single month — nearly a third of all buys in the entire seven-month log, crammed into the exact period when SOXL topped and rolled over. This is scaling-in's dark side. The same "add on every dip" logic that built a great base in the $40s-$70s became a liability when it kept adding into a falling market in the $200s. Scaling in doesn't protect you if you keep scaling into a downtrend — it just averages you into a bigger losing position, one tranche at a time.

The highest-priced buy of the entire run, $273.21, and that flurry of 14 June entries, tell the story: tranche-buying is a risk control only when the trend supports it. Buy every dip in an uptrend and you build a great position. Buy every dip in a topping market and you build a great problem.

What Every Investor Can Take From This

The scaling-in lesson generalizes cleanly to ordinary investing:

1. Averaging in beats betting it all on one entry. You almost never pick the exact bottom, and splitting purchases spreads that timing risk. This is the same logic behind dollar-cost averaging.
2. But averaging down into a downtrend is a different thing entirely. Adding to winners in an uptrend builds wealth; adding to losers in a downtrend builds exposure to a falling asset. They feel similar and end very differently.
3. Size your tranches so no single one can hurt you. The bot survived bad entries because each was just one tranche, not the whole account.

The Bottom Line

Across 50 buys and a price range from $42 to $273, the bot's tranche-based sizing was both its shield and, in June, its trap. Scaling in protected it from any single catastrophic entry and built the base for its biggest wins — but the same mechanism, applied to a topping market, quietly averaged it into its worst cluster of losses. The tool wasn't the problem. Using it in the wrong regime was.

Disclosure: Personal experiment, educational purposes only. Not financial advice. Leveraged ETFs carry substantial risk of loss.