TBLA — Taboola.com Ltd.

Four consecutive beats, then a miss. The streak ended on 5 August with $0.01 against $0.04 expected on a GAAP basis, and revenue 4.5% short. — Next: 4 November 2026

Zero open-market purchases in 180 days. The only figure that matters is Apollo's: 12,000,000 shares at $4.40 on 28 May, leaving 18,039,644 held, ten weeks ahead of a print that took the stock 27.5% lower in a session. Note that the aggregate insider feed lists this sale twice; the identical post-transaction holding on both lines confirms one transaction, not two. Everything else in the window is either a share award, an option exercise or tax withholding, none of which are purchases or sales of conviction.

Share count is falling, not rising. $64.2M of repurchases in the first half, roughly 14M shares at prevailing prices, partly offset by the 5,536,181-share executive grant in February. The last S-3 registration statement was filed on 28 February 2024.

Technical Setup

The trend broke in one session. On 4 August the stock closed at $5.2899; on 5 August it closed at $3.835 on 9,255,600 shares, a 27.5% fall on roughly three times the following day's volume. Price now sits 20.5% below its 20-day exponential average at $4.892, with the 50-day at $4.8617 stacked within six-tenths of a percent of it. That cluster is overhead resistance now, and it is a long way overhead. The relative strength index at 29.3 is oversold, which in a trend this damaged means very little on its own: oversold readings mark exhaustion in uptrends and merely describe the damage in downtrends. What matters more is the empty space beneath. The 5 August low at $3.5199 is the only tested floor between here and the 52-week low at $2.835, and the average true range of $0.3206 is 8.2% of the share price, so a single ordinary session covers most of the distance to that floor.

The regime is no alibi. A risk-on tape did not stop this stock falling 27.5% in a session, because the cause was company-specific and structural rather than macro. The mean analyst target of $5.79 predates the guidance cut and no revisions have landed yet, so treat it as a number about to move rather than a number to lean on.

Risk Synthesis

There is a useful distinction between a company that is cheap and a company whose share price has fallen. Taboola is both, but not for the same reason. The share price fell because revenue guidance was cut and the third quarter is now guided to shrink. The stock looks cheap because of a price/earnings ratio built on a courtroom settlement and a non-GAAP forward estimate. Those are two different statements, and conflating them is how value traps recruit. The honest way to own this is to underwrite the 2027 revenue line, not the multiple. Running the fundamentals through five threshold-based investor screens makes the split explicit, and it is worth doing twice. Fed the vendor's published figures, four of the five screens read bearish. Fed the corrected numbers — price-to-book of 1.11 rather than 1.53, and the 10-Q's $133.1M of cash against $72.0M of debt rather than the feed's roughly-net-zero — two flip to bullish and the panel moves to neutral. The deep-value tests now pass comfortably: balance sheet, book value, enterprise value against cash flow. The quality tests still fail, and they fail on the things that decide the next two years, namely a 29.7% gross margin against a 40% bar, an operating margin that is negative on a trailing basis, and return on equity of 11.2% against a 15% bar. Cheap on assets, weak on the business. That is the whole file in one line, and it is also why the grade is C+ rather than B or D.

Thesis

Price sits 20.5% below a 20-day average of $4.892 that now acts as resistance, with the 50-day stacked six-tenths of a percent below it. Beneath, there is almost nothing: the 5 August low at $3.5199 is the only tested floor before the 52-week low at $2.835, and with an average true range of $0.3206, roughly 8.2% of the share price, one ordinary session covers most of that gap. The map dies cleanly on a daily close above $4.16, which would take out the 5 August high and the $4.1202 that capped 6 August, and would say sellers have stopped setting the price. Be clear about what the first leg is worth: 1.4 to 1 is thin, and the whole reason to draw this is the second leg, where the 52-week low comes back into range if 4 November brings a third guidance cut. The counterweight is genuine and should temper any short: roughly $61M of net cash, an active buyback and no dilution machinery mean this does not go to zero on a bad quarter, it goes to a lower multiple.

Mindset Tip

When a company's press release leads with a metric you had never heard of before reading it, find the metric it replaced. Here the headline raised ex-TAC gross profit guidance, and the line it displaced was revenue, cut by up to $110M in the same document.