KO defensive profile vs SPY — downside vs upside capture (last ~3 years)
When SPY’s worst 10% of days rolled through the sample, Coca‑Cola didn’t merely lag less — it showed positive returns. Over ~3 years (749 trading days; 75 worst‑decile SPY days) KO averaged +0.0908% while SPY averaged −1.6682%, beating the market on 94.7% of those extreme selloffs. The downside capture of −0.118 versus an upside capture of 0.062 makes that asymmetry clear: KO cushions downturns rather than just being uniformly slow.
We examined minute bars resampled to daily closes and computed close‑to‑close returns, measuring capture as mean(KO)/mean(SPY) on SPY up and down days and testing KO−SPY differences on worst‑decile dates. The detailed statistics, charts, and significance tests below show the effect and support the thesis that Coca‑Cola’s defensive badge is real in this window.
For KO over the past ~3 years, does Coca-Cola actually earn its 'defensive' badge — on SPY's worst-decile down days does KO lose materially less than the market, and is its downside capture well below its upside capture, or is it just a sluggish low-beta name that lags in both directions? Thesis: KO's downside capture sits meaningfully below both 100% and its own upside capture, so it genuinely cushions selloffs rather than merely being slow, making the defensive label real rather than marketing.
How this was measured
Resampled KO and SPY minute bars to daily closes, computed close-to-close daily returns, and aligned by date. Evaluated upside capture as mean(KO return) / mean(SPY return) on SPY up-days and downside capture as the same ratio on SPY down-days. Defined SPY worst-decile days as the bottom 10% of SPY daily returns within the analyzed window and compared KO vs SPY returns on those dates, including a Welch one-sample t-test on the KO−SPY return difference (two-sided). The study window is capped to the most recent ~3 years available in the data.
The key numbers
Reading the numbers
Across 749 trading days, KO's upside capture is just 0.0616 (about 6.16%) while its downside capture is −0.1178 (−11.78%); on SPY's worst-decile days KO's capture is −0.0544 (−5.44%). In short, KO barely chases rallies and on aggregate cushions or even bucks selloffs.
The charts
The box plot compares KO and SPY returns on the 75 SPY worst-decile days: KO's mean is +0.0009 while SPY's mean is −0.0167, so KO averages a small positive return when SPY is deeply negative. KO's range runs from −4.38% to +3.41%, whereas SPY's worst-decile range is entirely negative (−5.75% to −0.97%), so you'll notice KO often posts non-negative outcomes on days when SPY never does. That visual gap — KO centered near zero while SPY is clearly below zero — is direct evidence that KO cushions selloffs.
The capture-ratio bars make the asymmetry plain: upside capture is only 0.0616 (6.16%), downside capture is −0.1178 (−11.78%), and worst-decile capture sits at −0.0544 (−5.44%). Look at how downside capture is not just below 100% but is negative and materially farther from the upside bar — KO participates minimally in rallies and shows a clear asymmetric downside behavior, consistent with a genuine defensive profile rather than simple slowness.
KO vs SPY capture summary
| metric | value | window_start | window_end |
|---|---|---|---|
| Sample days | 749 | 2023-07-03 | 2026-06-30 |
| SPY up-days (N) | 418 | ||
| SPY down-days (N) | 330 | ||
| SPY worst-decile days (N) | 75 | ||
| Mean KO on SPY up-days | 0.0004 | ||
| Mean SPY up-days | 0.0067 | ||
| Upside capture (KO/SPY) | 0.0616 | ||
| Mean KO on SPY down-days | 0.0008 | ||
| Mean SPY down-days | -0.0067 | ||
| Downside capture (KO/SPY) | -0.1178 | ||
| Mean KO on SPY worst-decile | 0.0009 | ||
| Mean SPY worst-decile | -0.0167 | ||
| Worst-decile capture (KO/SPY) | -0.0544 | ||
| KO outperform rate on worst-decile | 0.9467 | ||
| Mean (KO − SPY) on worst-decile | 0.0176 | ||
| Daily beta (slope KO~SPY) | 0.0643 |
The takeaway
Yes — over the past ~3 years KO does behave like a genuine defensive stock. On SPY's worst-decile days (N=75) KO actually averaged +0.0908% while SPY averaged -1.6682%, so KO beat the market by about 1.76 percentage points on those extreme down days. KO outperformed SPY on 94.7% of those worst-decile dates, and the KO−SPY gap is highly significant (t = 10.64, p ≈ 1.44e-16), meaning this is overwhelmingly unlikely to be just sampling noise. The capture numbers underline the asymmetry: downside capture ≈ -0.118 versus upside capture ≈ 0.0616, and the low daily beta (~0.064) shows KO barely tracks SPY on up-days while holding up in selloffs. Practical takeaway: with 749 trading days and a large, statistically clear worst-decile effect, Coca‑Cola’s defensive badge is real here — it cushions major market drops rather than just trailing rallies.
The fine print
- Worst-decile is defined in-sample for 2023-07-03 to 2026-06-30; different windows change which days qualify.
- Capture ratios use simple arithmetic mean returns by day sign; geometric/CAGR definitions can differ.
- Close-to-close returns omit intraday paths and gap opens, which can dominate extreme days.
- Analysis is limited to ~3 years; a longer history could tighten or shift the measured asymmetry.