They Doubled Size and the Loss Doubled Too: Week Five of an Automated Options Portfolio

Last Monday the team decided to double position sizes across the portfolio, and the verdict came fast: it was the worst possible week to scale up. Only two trades opened after that decision and both closed as losers, and both were the largest single losses to date.

The Option Alpha portfolio is a menu of ready-made strategies that can be mixed into a custom basket, and the team runs its own selection, the Power 7, live. The idea is to combine seven uncorrelated strategies across symbols and expirations so that one can offset another. The firm committed 50,000 dollars of its own capital to the setup, with every trade fully automated and no manual overrides allowed.

A stricter-entry version of the same seven strategies, the Lean Seven, is also shared as a community template. It targets smaller accounts with lower drawdowns and reduced size. Expectations were set firmly: the portfolio demands at least a three-month commitment, drawdowns of one to two thousand dollars are routine, and more than 48 percent of trades end in the red.

The numbers show how harsh the week was: after booking a 1,438-dollar profit last month, the portfolio slid to minus 4,000 dollars this month. One of the two trades came from an afternoon-triggered setup, the other was a gamma scalper stopped at a loss. A touch-exit order that failed to fill could have kept the second loss smaller, as noted on screen.

The core lesson is about sizing timing: weeks of gains built at small size were erased in the single week size was doubled, with extra damage on top. The team refuses to dramatize it and frames the loss as part of a long-term plan to be accepted and traded through.

The roundtable added personal balance sheets: the host admitted he had doubled size in his own account and took the same haircut. Amanda overtraded manually last week and early this week, drained her bot allocation, and had to pause the portfolio bot temporarily. Steve argued against five-week panic with a baseball analogy: a .300 hitter who goes 0-for-10 does not throw away the bat.

The losing trades were also placed inside their backtested context: the strategy behind the big daily loss wins about 85 percent of the time over the long run, so a 2,388-dollar loss at double size sits within its normal range. Its worst historical single loss stands near 3,000 dollars. An unpleasant print, in other words, but an ordinary row in the model's own history.

Next week the basket grows to ten: four new variations join the visible list of ten strategies, several sourced from the community, and the Power 10 will be born. Setups found by community member Aheem among more than 1.3 million backtests earned a public shout-out. An incoming SPY income strategy with a win rate near 90 percent is expected to add frequency and activity.

The week's most tangible platform upgrade is analytical: every backtest now shows VIX and market-event breakdowns alongside reward-risk breakdowns. In the flat-flyer structure, the last day of the month showed the lowest win rate, plausibly linked to large funds rebalancing. On the reward-risk side, trades taken at very low ratios were shown to drag overall results.

A batch of small functional fixes shipped for bots and backtests: bot scanners generated from a backtest now run only on the strategy's trading days, and combined portfolios gain yearly totals next to monthly splits. Position lists default to newest-first order with weekday labels. Each change aims to remove one-by-one backtest repetition from strategy selection.

The reward-risk ratio will soon become a range with both a floor and a ceiling, the team announced. The evidence comes from the new screen: trades with ratios above 1,600 lost almost without exception, usually a sign the market behaved oddly that day. In one host experiment a single filter lifted profit slightly while cutting drawdown by about 30 percent and removing forty trades.

The Q&A opened with a transparency promise: a live results tab will show the bot's open and closed positions in real time. A strategy-submission screen is coming with minimum return-per-drawdown rules measured from early 2023, slippage floors, and expiration-risk constraints, so community portfolios can be cloned. Double size stays manual and takes about two minutes; it will not be baked into the shared template.

The profit-target question got a backtested answer: adding a 75 percent target to the afternoon setup raised the win rate from 84.8 to 88 percent but cut profit by 6,000 dollars. A 50 percent target erased more than half the profit while drawdown stayed identical. No technical indicators will join the analysis screens, for data-availability and readability reasons.

The closing round collected practical notes: the Power 7 template remains 1x and only scanner days changed; fees and commissions sit on the near-term roadmap. Whether Power 10 ships as a separate bot or an upgrade is undecided. Small accounts were pointed at a four-strategy Lean selection with VIX stops built through monitors; undefined-risk strategies, futures, and calendar spreads stay out of scope.