Tickbird
Back to the track record

Wed, Sep 30, 2026 session

Today's pick

High conviction

KMX

CarMax Inc

-6.91%S&P 500 -0.50%lagged the market
3 of 3 AIs said yes
Consensus score79.3
Confidence75.3
Post on XPost to Reddit

What the AIs concluded

The strongest near-term catalyst is CarMax's September 29 earnings beat, including 19.5% revenue growth, an approximately 81% EPS increase, and nearly 19% same-store sales growth. The reported plan to resume share buybacks and a recent Stephens price-target increase add support. However, much of the good news may already be reflected after the stock's sharp move, while cautious analyst ratings and below-market consensus targets create downside risk. The next session's bias is bullish but likely volatile.

Catalysts

  • CarMax reported fiscal Q2 2027 results on September 29, 2026: revenue rose 19.5% year over year to $7.88 billion and EPS increased about 81% to $1.16, substantially above the reported $0.73 consensus estimate.
  • Same-store sales increased 18.9% year over year, indicating strong near-term vehicle demand and supporting the positive earnings reaction.
  • CarMax posted strong earnings with revenue up 19.5% and EPS up 81% year over year.
  • CarMax just reported an 81% EPS surge and 19% same-store sales growth, far exceeding analyst expectations, which typically drives continued buying.
  • Stephens raised its price target from $66 to $74, suggesting analysts see more upside from current levels near $59.

Risks flagged

  • The stock already rose roughly 4.7% on the latest session after the earnings release, increasing the risk of profit-taking or a sell-the-news reaction.
  • Analyst sentiment remains cautious overall: the reported consensus rating is Reduce, with average price targets near $51-$53, below the supplied current price of $59.23.
  • Much of the good news may already be priced in after the recent sharp move.
  • The stock already jumped nearly 5% today on high volume, so much of the good news may already be priced in, limiting further gains.
  • The 20-day return is still negative (-3.4%), suggesting the stock was weak before earnings and some investors may use this bounce to exit.

How it turned out

-6.91%S&P 500 -0.50%lagged the market

Open $59.14 → close $55.11, with 0.1% simulated trading costs deducted.

Google or Yahoo may show a bigger number for the same day. They measure from the previous close, which includes the overnight move that had already happened before you read the pick. We only count what you could have captured: open to close, on the day.

The bar we set that morning

These thresholds were fixed before the analysis ran and are sealed into the fingerprint below, so they cannot be softened after the fact to make a call look better.

Score required
70.0 (68.0 base, raised 2 for a mixed market)
Confidence required
62%

Publication fingerprint

This analysis is sealed with a SHA-256 fingerprint computed the moment it was published. Any later edit would break it, which is what stops us from quietly rewriting history.

1b9af6c901f83477c06c6de6ae7a114c9adcfc5a0b6c16f8fafd8b44cf8b3183

Produced by algorithm v0.6.0, which is sealed inside the record above. If a newer version is running today, this call still names the rules it was actually made under — see the version history.

Do not take our word for it. The exact bytes that were sealed are public, so you can recompute the fingerprint yourself and compare:

curl -s https://tickbird.ai/api/public/signals/e161c515-62fd-4c0d-b331-7f1c2f29e70e/snapshot | sha256sum

On macOS, use shasum -a 256 instead.

Save that file today and run the same command in a year. If we had quietly edited a single character, the two fingerprints would no longer agree.

Open the sealed record

Anchored in Bitcoin

A fingerprint on our own server only proves the record still matches the number we published. It cannot prove the number is the original one, because we run the server. So the fingerprint is also submitted to OpenTimestamps, which writes it into the Bitcoin blockchain — a ledger nobody here controls.

Attested by Bitcoin block 969358. That block proves this analysis existed, word for word, before it was mined.

Check it yourself with the reference client, which talks to Bitcoin rather than to us:

pip install opentimestamps-client
curl -s https://tickbird.ai/api/public/signals/e161c515-62fd-4c0d-b331-7f1c2f29e70e/snapshot -o snapshot.json
curl -s https://tickbird.ai/api/public/signals/e161c515-62fd-4c0d-b331-7f1c2f29e70e/snapshot.ots -o snapshot.json.ots
ots verify snapshot.json.ots

The date it attests is when the fingerprint was submitted, which for older analyses can be later than the session itself.

This is not investment advice. Tickbird is an automated experiment. It publishes a general analysis, not a personal recommendation, and it makes no promise about any outcome. Past results do not predict future ones. Never risk money you cannot afford to lose.