What I traded: stocks, leveraged products and the FCN

I continued holding Ping An and received its dividend. Its share-price recovery and dividend helped the month. At the same time, I actively traded the AI and memory theme through several different instruments and entered a Micron–SanDisk fixed coupon note (FCN). The note’s month-end valuation decline became the largest drag; its eventual maturity outcome was still unresolved.

  • SKDD, 07709 and SKUU: I traded the inverse SKDD product and the leveraged long SK hynix products 07709 and SKUU. These were different instruments, but their risks were connected to the same memory theme.
  • TSM, KLAC and LRCX: I sold the TSM position carried into July. Two KLAC stock trading rounds ended in losses. I also bought LRCX and sold it again within roughly a week.
  • SNDQ: I made repeated trades and still held a position at month-end. SGOV was established for liquidity management. Holding a liquidity instrument did not remove the risks in the other products.

What I traded: options and individual events

  • CRCL and MSFT: I sold puts and later bought them back at lower option prices. Some Laopu Gold and Pop Mart short puts expired unexercised; they were not assigned. A separate Pop Mart short-put position remained open at month-end. The distinction matters: one contract expiring does not mean all exposure to that company has ended.
  • Google and Nokia: the Google share trade was profitable. Nokia combined shares with short calls; repurchasing the calls and paying costs left the combined trade at a loss even though the shares were approximately flat.
  • IPOs and Momenta: two IPO applications were refunded in full but still incurred fees. I subsequently bought Momenta through grey-market / spot trading and sold in two stages. This was a market trade after allocation attempts, not a profitable IPO allotment.
Understanding a company and choosing how to take the risk are different decisions.

My reflections · edited from the review

The research became more detailed, but so did the collection of products I used to express it. Equipment shares, leveraged ETFs, inverse ETFs and the FCN looked like separate decisions. Much of the portfolio still depended on AI spending and the memory cycle. I needed to count the underlying risks, not just the number of tickers.

Ping An’s recovery and several successful short puts could not prevent the FCN valuation change from dominating the month. I cannot explain that away by saying the industry research was sound. Entry price, instrument choice, concentration and the ability to exit were part of the investment decision too.

The options review also needed precision: expiration without assignment is not the same as taking delivery of shares, and a newly opened contract is not the same trade as one that just expired. Getting the sequence right is necessary before judging whether my decisions were good.

My AI research partner’s suggestions

These priorities were proposed for the following month. They should not be read as proof that August’s trading already followed them.

  • For each thesis, identify the core position and any specific hedge. Review additional products as overlapping exposure, and write the catalyst, invalidation condition and exit / re-entry rule before trading.
  • Review the FCN through its weakest underlying, strike and barrier conditions, observation dates and maturity scenarios. A coupon alone is not a risk assessment, and an inverse ETF does not automatically match the note’s exposure.
  • Check short puts against the liquidity needed if they are assigned together. Record long-term holdings, options, event trades, IPOs and hedges separately, then bring them back together to assess portfolio risk and costs.

The reported result

The estimated Modified Dietz return was −1.08%, after adjusting for external cash-flow timing. The FCN’s mark-to-market decline belongs in that result even though its maturity outcome was not yet known. A hypothetical return excluding the note cannot replace the actual month.