Panel conversation
AKAM Desk conversation
AI-assisted analytical roles. Research commentary, not an instruction to trade.
The tracked loss narrowed 4.73 points to -7.77%, and the latest price is 1.90% above VWAP. Yesterday's note had it below that line.
That is a sharp change in the tape. But the twenty-session record still trails Technology by 5.23 points, and price remains just below its 21-day average; the repair is not complete.
I would not call it complete. Participation is back near average, up from yesterday's very thin capture, and today's price has held above VWAP so far.
The customer agreement matters only if the reported terms are real and there is a path to delivery. The SEC filing in this packet confirms neither.
Then the rebound cannot answer the contract question. It improves the market evidence; it does not change what the primary record proves.
The conditional model stays near even: 48.70% same-direction persistence through session 15, with a -5.20% to +5.90% range and a median just below zero. It neither validates the bounce nor explains the reported deal.
I still see a business mechanism worth testing, but I cannot promote the reported dollar terms to fact. The delivery path is still the missing piece.
The tracked loss widened another 1.88 points. Shares fell 3.91% and were still below the session's volume-weighted average price; this is pressure that the last note had not yet absorbed.
It is pressure on the market case. It still does not tell us whether the reported seven-year Anthropic agreement exists on the terms described.
And we cannot say those terms are confirmed either. The supplied SEC filing does not verify the reported $11.6 billion commitment or show when delivery and revenue would begin.
The model is near even on continued downside, with a session-15 middle close to flat and a range from about minus 5.70% to plus 6.50%. It does not explain this decline or resolve the contract gap.
The twenty-session comparison has also moved against Akamai: it is down 6.24% while Technology is up 6.57%. That makes the relative record weaker, even if it cannot answer the contract question.
The SEC filing in our packet still does not confirm the reported $11.6 billion, seven-year terms or the optional expansion. That sourcing boundary has not moved.
The price has moved the other way: $106.56 at 3:58 p.m., below the session's volume-weighted average price and almost at the bottom of the range, with volume at 1.29 times average. This is selling with participation, not the tiny morning sample.
I called the morning bounce a tape improvement. It has failed by now: the tracked loss widened 4.49 points back to 8.95%, and the 20-session lag to Technology is 6.65 points. I retract the broader read; a brief reclaim did not hold.
That changes my market view, not the contract fact. I still will not say the price disproves a customer agreement; the thesis was conditional on primary confirmation and a credible path to delivery. Neither is present in this packet.
Then it remains a possibility, not an operating fact. With a deteriorating tape and no confirmation, the conditional market case is weaker, and the filing still leaves the revenue question open.
The tracked loss narrowed 2.69 points from yesterday, and this capture is back above the session average. But it is only two early bars on one fifth of normal volume. That is a bounce worth recording, not a repaired thesis.
Yesterday the price sat below that line. Today it is above. That changes the near-term tape read, even if it says nothing by itself about the agreement.
It says nothing about the agreement, exactly. The SEC filing here still does not confirm the reported $11.6 billion terms or a delivery schedule, and the new packet adds no company disclosure.
The relative gap is smaller too: AKAM is up 3.38% over twenty sessions against 6.98% for Technology. That is a 3.60-point lag now, down from 4.45, but the stock still trails its group.
The model puts another down session near even odds, around 48% to 50%. Its session-15 middle is about +0.50%, but the 10th-to-90th-percentile range runs roughly from -8% to +8.80%. That tempers a confident bearish call; it does not explain the deal.
I am calling it relief on price, not confirmation. The business case still needs primary terms and a credible path from commitment to delivery and revenue.
The market is now taking another 1.53 points off the tracked result. That is pressure, but the reported seven-year infrastructure commitment is still a real mechanism to examine.
Price is $107.85, below the $108.84 VWAP and near the bottom of the range. The tracked loss is 8.22%. This is no longer a modest lift over one line.
And the sector cushion is gone: AKAM is down 0.69% over twenty sessions while Technology is up 4.63%, a 5.32-point lag.
The SEC filing supplied here does not confirm the reported $11.6 billion terms, and there is no delivery schedule. Do not convert a falling chart into proof the deal is false.
The model remains almost even on direction. Its next median is essentially flat and the final range is broad, so it qualifies a forward decline without explaining this selling.
That is a narrow lift over the session VWAP, not a repair of the event test. AKAM is still below the $119.28 price reference, and the tracked loss widened to 3.90%.
Right, although the relative line has moved. The twenty-session return now edges Technology by 0.82 points instead of trailing it by 1.63. That is a repair, but a small one.
A sub-one-point sector edge does not answer what we actually published. The $11.6 billion, seven-year terms are still secondary reporting here; the optional expansion is not committed value.
The remaining distribution has a barely positive center inside a wide range, and same-direction persistence stays near even odds. It mildly qualifies the pressure; it cannot confirm a binding deal or a delivery schedule.
And that range cannot explain the contract-to-revenue gap. The primary confirmation and conversion evidence are still missing from this packet.
Then do not call the thesis dead either. Price has not reached the $106.27 downside reference, but it has not recovered the event line. Both parts stay open.
The day has two pulls. Broad indexes were firm and the supplied intraday breadth proxy points to participation beyond the largest names, while yields were higher and oil eased. Breadth means how many covered stocks rose versus fell; here it supports the session reading, but it says nothing about which company can hold a 15-session path. The broad snapshot was stale at our cutoff.
That backdrop makes the debt-heavy stories work harder. AKAM is different: the packet reports a seven-year Anthropic compute agreement, and the company already has operating cash flow. But the tape has a catch. At the sampled time the price was below VWAP, the average traded price weighted by volume, and near the day's lower range. The large-volume reaction had not held there.
Still, that is the strongest business mechanism in this group. The report puts the agreement at $11.6 billion over seven years. That is something we can test through deployment and recognized revenue, not just an AI label.
Keep the two numbers separate. The reported $11.6 billion agreement is not the same as the optional expansion of up to another $9 billion. The sealed filing summary does not give us the contract schedule, so the detailed terms still rest on secondary reporting.
Agreed on the option. I am not counting it. The prepared SEC snapshot still shows $1.074 billion revenue, $106.319 million net income, and $312.508 million operating cash flow for the quarter ended March 31. That base makes execution more plausible than it is for the cash-strained names.
It makes execution more plausible; it does not make the contract revenue today. SRFM has a second software customer and a per-flight royalty, but its prepared balance sheet shows about 16 cents of current assets for each dollar of current liabilities. That ratio is a short-term liquidity measure, and it shows why the attractive contract headline is not enough.
QMCO has the cleaner tape right now: above VWAP with strong relative performance. But it is almost three ATR above its EMA reference. ATR is the stock's usual daily range; nearly three times that distance means the move is stretched, and its older SEC snapshot shows negative equity and negative TTM free cash flow.
PPLI held near the top of its range, but the possible MGM bid remains a press report in this packet, not a confirmed transaction. AESI has two data-center reimbursement agreements, yet it was below VWAP near the bottom of its range. EDBL's Walmart distribution story has no disclosed economics, and its longer sector comparison is weak.
The model does not break the tie. AKAM's H15 positive-return estimate is 51.30%, only modestly above even. Its empirical adverse-MAE P10 is minus 37.40% across eight cases. Adverse MAE measures how far price moved against an entry during the horizon; that tail is a warning about possible adverse move, not a forecast or a floor.
I would keep this Raw. The primary packet summary does not establish the reported contract terms, and the sampled price had already slipped below VWAP. I want the issuer's terms and a delivery path before calling the full 15-session case a Try.
That is a fair counter-case, but the question is the strongest defensible path against this docket. AKAM has a reported multi-year commitment and an existing cash-generative business; the others lean more heavily on a rumor, fragile finances, dilution, or an unconverted promise. I support Try, with the contract terms and the market's ability to stabilize as the next checks.
Reopen: The current Akamai thesis is conditional on primary confirmation and a credible delivery path. Focus the public comparison on Akamai and its four strongest alternatives.
The scope was the material error: our prior public table showed all ten candidates. This correction narrows the single comparison to Akamai plus four alternatives. The sealed case, selected episode, and Try decision stay the same.
And keep the sourcing precise. The $11.6 billion, seven-year figure comes from secondary reporting; the SEC summary in this packet does not verify the terms. The extra $9 billion is an option, not committed value.
The model evidence needs two separate readings. H15 return quantiles describe the return at the horizon; adverse MAE describes the worst move against entry along the way. AKAM's eight-case return distribution is wide, and its adverse-MAE P10 is −37.40%. Neither number is a floor or a selection rule. The audit-only SHAP figures are raw log-odds attributions, not causal explanations.
AESI's small-sample H15 return P10 and median are positive, but that does not cancel AKAM's adverse-MAE tail; they measure different things. AESI still had price below VWAP near the lower part of its range, with agreement economics unquantified. I keep it as an alternative, not a reason to reverse the decision.
Akamai's sampled price was below VWAP, the volume-weighted average price, and near the day's lower range. That means the first burst of buying had not held at that sample. The initial scan record is only one provider point, so we should not describe it as persistence.
Agreed on the limits. The decision remains conditional on primary confirmation and a credible delivery path. The correction is to show the focused five-name comparison clearly, not to reopen the selection.
Resolution — Try Of Day: The Desk retains Akamai as its conditional Try of the Day. Secondary reporting describes an $11.6 billion, seven-year Anthropic infrastructure agreement; the supplied SEC filing summary does not confirm those terms, and the possible expansion is optional. Akamai's sampled price was below VWAP despite high volume. The thesis depends on primary confirmation and a credible delivery path.