> ## Content Index
> Fetch the complete content index at: https://www.constellationstocks.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# October 2026 Recap: Down 7.8%, and a New One-Month Rule
- URL: https://www.constellationstocks.com/blog/october-2026-recap/
- Published: 2026-11-05T17:00:00.000Z
- Updated: 2026-11-07T04:47:48.000Z
- Description: October lost 7.84% against −0.93% for the S&P 500. One old position caused most of it, and my own trade data led to a new rule.
- Author: Iskan Rakhmatov
- Tags: Track Record

October closed at −7.84%, against −0.93% for the S&P 500\. The year is now down 34.47%, while the S&P 500 is up 10.74%. Most of the month's loss came from one position I had held since June, and that position led me to a new rule.

## The year so far

| Month            | Strategy    | S&P 500     |
| ---------------- | ----------- | ----------- |
| January          | +8.45%      | +1.37%      |
| February         | −1.72%      | −0.87%      |
| March            | −1.44%      | −5.09%      |
| April            | +11.02%     | +10.42%     |
| May              | +11.51%     | +5.15%      |
| June             | −20.99%     | −1.06%      |
| July             | −20.08%     | −0.13%      |
| August           | −15.68%     | +2.62%      |
| September        | +2.69%      | −0.45%      |
| October          | −7.84%      | −0.93%      |
| **Year to date** | **−34.47%** | **+10.74%** |

From the May high, the drawdown is now 49.6%, the deepest point so far. All figures are time-weighted returns from monthly brokerage statements, so deposits and withdrawals do not affect them. 2026 figures are unaudited. S&P 500 figures are price returns.

## What went wrong

**An old position.** The largest loss of the month came from Dropbox (DBX) puts that expired worthless on October 16\. They were the long leg of a bearish put spread opened in June. The short leg had already expired worthless, and I kept the long puts because I saw Dropbox as a struggling, ex-growth business. The view on the business may still be right. The timing was not: the market rotated into value and "boring" stocks, and DBX re-rated higher.

**Bullish bets on AI infrastructure, ad-driven businesses and crypto-linked names.** I went into October leaning bullish, and two things went against these positions. First, the market grew cautious about whether neoclouds, the new AI cloud providers that borrow heavily to build data centers, can repay their debts, while new data center launches were delayed. That hit Applied Digital (APLD), which builds and leases AI data centers. Second, Google's quarterly report showed how fragile the classic traffic-generation ad business is, as search leans further into AI. That weighed on Pinterest (PINS), a turnaround story that depends on ad revenue. My long calls on APLD, PINS, Twilio (TWLO) and Hyperliquid Strategies (PURR) closed at losses.

## What worked

- **Fabrinet (FN), bearish put spread.** An overvalued AI-related manufacturer in Thailand with a light moat.
- **MP Materials (MP), bearish puts.** I expected interest in onshoring rare earth extraction to cool.
- **AppLovin (APP), bearish put.** An ex-growth story.
- **Microsoft (MSFT), long calls.** The position from September, on a rebound after the Azure sell-off, closed with a gain.
- **Penguin Solutions (PENG), calls.** A net gain across several closes during the month.

October was the mirror image of September. In September, my bets against AI stocks lost, because AI stocks held up better than the rest of the market. In October, my bets on AI and ad-driven stocks lost, while my bets against overvalued names like Fabrinet worked.

## The lesson: close what has not worked within a month

The Dropbox puts made me look at how long I hold positions. I grouped every trade I closed from July to October by how long it was open:

| Held              | Positions | Winners | Average result per trade |
| ----------------- | --------- | ------- | ------------------------ |
| 30 days or less   | 40        | 52%     | About break-even         |
| More than 30 days | 19        | 26%     | A clear loss             |

Trades held longer than a month won only one time in four, and caused almost 90% of the losses in the period. The worst of them, including Dropbox, were bets that the market would come around to my view if I waited. It did not.

So I am adding a rule to the ones I set in August: **a position that has not worked within a month gets closed.** The data covers four months, so I will keep checking it as more trades close.

## What I did at the end of the month

On October 19 and October 28 to 29, I closed most of the book: seven calendar spreads and three vertical spreads. It was both taking profits on the trades that had worked, such as Microsoft, Fabrinet and MP Materials, and clearing out bullish positions before changing my view. At the end of October I held positions in 8 stocks and ETFs, well under my limit of 14.

## How I am positioned now

I am going into November prepared for several outcomes, not one forecast. Every week I map the catalysts due in the next one to three months, with the odds that prediction markets, options prices and my own analysis give each outcome, and what each outcome would do to the market and to my positions.

For the next 30 days, these are the events that matter most:

- **November 10, US inflation.** Markets give a 43% chance that core CPI comes in hot, above 0.2% for the month. A cool print, below 0.1% (18%), would be worth about +2.4% for the S&P 500.
- **November 10, China's rare earth export controls.** The suspension expires. The likeliest outcome is an extension (55%). A full reinstatement (10%) is the main risk for my MP Materials puts.
- **November 17, Nvidia results.** A beat and raise is the base case (70%). The AI hardware build-out rolling over has low odds (6%) but would cost the S&P 500 about 3.7%.
- **Late November, the Anthropic IPO,** and whether the market accepts a valuation of $1.5 trillion or more.
- **The Fed.** Markets price a 78% chance of a 0.25% rate increase on December 9, and 35% that the hiking cycle keeps going after that.
- **November 30, the Israel Iran ceasefire.** If it breaks (38%), the market is back in a regime of sideways moves or even a decline, the costliest single outcome on the map.

The book is small and built around these dates: puts on MP Materials and Salesforce, calls on Penguin Solutions and Himax, and a call spread on small caps (IWM). Each position expires after the event it depends on.

## The full record

Every month since January 2025 is on the [Track Record](https://www.constellationstocks.com/track-record/) page. In 2025 the same method returned 97.71%, verified by a U.S. licensed CPA. 2026 shows how it fails when positions are held too long in a market that changes fast. The rules I added in August, and the one-month rule from this report, are my answer to that.

I will publish a recap like this one every month, good or bad. Last month's is here: [September 2026 Recap](https://www.constellationstocks.com/blog/september-2026-recap/). If you want to see how the Method works on a live trade, [book a free walkthrough](https://www.constellationstocks.com/consultation/).

*Past performance is not indicative of future results. This post is educational and is not investment advice.*