Investment Journal / Market notes
The week Alphabet fell 12.6% in a day — assessed with indicators set in advance, not price alone
A week when big tech swung hard on a single earnings report. The decline was concentrated in a few large tech names, and the US investment-grade corporate spread held between 0.86 and 0.91 percentage points. I didn't sell on price alone; I judged by indicators I'd set in advance — favorable for gold, unfavorable for copper.
Now
Big tech fell after Alphabet's earnings
Alphabet reported third-quarter results after the close on October 28. Revenue rose 25% year over year to $127.6 billion. But capital expenditures (spending on data centers and servers) more than doubled from a year earlier to $51.4 billion, and free cash flow (operating cash flow minus capex) was −$9.1 billion.
The stock fell 12.6% the next day. The same day, enterprise workflow software company ServiceNow fell 27.0%.The indexes fell, but more sectors rose than fell
For the week, the Nasdaq fell 2.0% and the S&P 500 fell 0.6%. The small-cap Russell 2000, by contrast, rose 1.3%.
Looking at sector exchange-traded funds (ETFs), health care (XLV) rose 3.1%, aerospace & defense (ITA) 3.7% and utilities (XLU, power and gas companies) 2.0%.
In other words, the decline was concentrated in a few of the largest tech stocks, not spread across the market.The KOSPI crossed 6,000 for the first time, then pulled back
On October 28 the KOSPI rose 6.6% in a day and crossed 6,000 for the first time (6,012). SK hynix rose 10.9% that day. On Friday the KOSPI fell 3.5% to close at 5,731. For the week, the KOSPI rose 3.3% and the KOSDAQ fell 3.6%.
Yields near 5%; gold and silver were weak
The US 10-year Treasury yield rose as high as 5.05% on October 28 and ended the week at 4.93%. The 30-year yield is 5.31%. Gold fell 2.7% for the week and silver 8.6%. Oil (WTI) stayed above $100 a barrel, at $104.
Corporate bond spreads barely moved
A corporate bond spread is the extra yield a corporate bond pays over a Treasury of the same maturity. It shows how much default risk lenders are pricing in.
The US investment-grade corporate spread stayed between 0.86 and 0.91 percentage points all week (0.89 points on October 29). Stocks were shaken, but the bond market barely reacted.
Outlook — what I'm watching, and what would change my mind
Will other big tech companies cut capex?
- If so
- If other big tech companies announce cuts to data center and server spending plans in upcoming earnings reports, fewer new servers get switched on, and the outlook for the power they would use falls with it. That would be the first crack in the thesis that AI data centers will use ever more power — and I'd re-examine the thesis behind my growth assets (AI power and the power grid).
- If not
- If plans hold, I'll read this drop as the market repricing cash flow, not investment plans.
Will corporate bond spreads widen as stocks fall?
- If so
- If the investment-grade spread rises above 1 percentage point alongside falling stocks and stays there for more than a month, I'll read it as an earnings disappointment turning into a credit problem. I'd then first check whether I can ride out a drop in everything at once (cash and time horizon).
- If not
- If they stay near today's level (around 0.9 points), I'll read it as a reaction to a few large companies' earnings.
The indicators behind my sell rule for gold and copper — compared with three months earlier
- Current reading
- Under the sell rule I set on October 28, I compared each indicator with its level three months earlier.
For gold I use 10-year breakeven inflation (the average inflation rate the market expects over the next 10 years). When inflation looks set to rise, demand for gold as a store of value tends to rise, so I count that as favorable. It's 2.34% now, above three months ago (2.27% on July 30) — favorable.
For copper I use the gold/silver ratio. It's 79.4 now, above three months ago (69.7) — unfavorable. More than half of silver demand is industrial (electronics, solar), while most gold demand is jewelry, investment and central bank holdings. So when silver is much weaker than gold, I read it as a sign that factory demand for metals is falling. Copper is also an industrial metal, used in wiring and motors, so I use this ratio as its indicator. - Next
- I trim assets with unfavorable readings back to my preset weight limits, even at a loss — which is why I decided to sell part of my copper by October 30 (see related decisions below). For assets with favorable readings, I wait until my quarterly review on December 31. Until then I'll record both numbers every week.
Will the 10-year yield go back above 5%?
- If so
- If it climbs back above and stays there, rate-sensitive assets such as gold and dividend stocks come under pressure together. Different holdings, same risk (rates) — I count them as one group.
- If not
- If it falls below 4.9%, one cause of this week's weakness in gold and silver fades.
My view — read through my principles
This week showed in prices why I split the portfolio into stable and growth assets (a core-satellite strategy). On the growth side, an AI platform company swung hard on one earnings report, while essential-service sectors (health care, utilities) rose the same day. Still, one week's price moves aren't enough for me to judge whether the strategy worked.
I won't sell Alphabet solely because its stock fell 12.6%. I have only three reasons to sell: exceeding a weight limit, a broken thesis, or swapping into a better opportunity. So I separate the causes. If the problem is AI eating into search and advertising, the thesis is at stake. If it's capex growing faster than cash flow, it's a price question — as long as that spending is turning into revenue.
The numbers I could confirm this time (cloud revenue up 103%, capex doubled, negative free cash flow) all point to the latter. I found no figures showing search or ad revenue shrinking. So for now I treat it as a price question. Next quarter I'll check search and ad revenue separately, and if cloud revenue growth turns down, I'll look at it again as a thesis problem.
The number I watch most closely is the gold/silver ratio. On October 28 I decided that if this ratio is higher than three months earlier, copper reads unfavorable and gets trimmed back to its weight limit, even at a loss. This week the ratio rose from 74.6 to 79.4, tilting further that way. This is the first time I'm actually using the rule of basing my decision on an indicator set in advance rather than on whether I've broken even.