By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel consensus is that relying on a 'dividend run' strategy for Amgen (AMGN) is unreliable and risky, given the small sample size, market beta contamination, and the looming patent cliffs on key drugs. The upcoming ex-dividend date (02/13/26) may not result in a significant pre-ex run-up, and even if it does, the net, risk-adjusted payoff is unlikely to justify the strategy.
Risk: The 18-24 month patent cliff on key drugs like Enbrel and Otezla, which could compress the company's valuation and offset any short-term gains from a dividend run.
Opportunity: The potential re-rating of Amgen's valuation if obesity drug MariTide's clinical trials yield positive surprises.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
This morning a "Potential Dividend Run Alert" went out for Amgen Inc (NASD: AMGN), at our DividendChannel.com Dividend Alerts service (a free email alerts feature). Let's look at the situation in greater detail, shall we?
First of all, what is a "Dividend Run" anyway? This is an interesting concept which we first learned about at a past ValueForum conference. And to best explain the concept, we need to start with the expected behavior of a stock on its *ex-dividend* date.
For anyone unfamiliar with the term, the *ex-dividend date* marks the trading day when any buyer of the stock is no longer entitled to the referenced dividend — in other words, to be eligible to receive the dividend in question, one would have had to purchase their shares *before* the ex-dividend date.
All else equal, the stock price would be expected to *drop by the dividend amount* on that *ex-*date (remember, that's "all else equal" and naturally other factors will drive stocks higher/lower on any given day). But think about it: if a buyer is entitled to a 2.52 dividend *before* ex-date, but no longer entitled to that amount *on or after* ex-date, then this drop makes perfect sense! Because if the shares *didn't* drop by that same 2.52 the next day, then *effectively*, buyers would effectively be paying 2.52 *more* for the same share of stock.
But now think about *this:* if a stock is expected to *drop* by the dividend amount (all else equal) *on* ex-date, then in turn, shouldn't that stock be expected to *rise* sometime *ahead* of a dividend? After all, if a dividend-paying stock didn't ever rise and only fell on each and every ex-date, then eventually after enough dividend payments those shares would have fallen to zero. And *that* wouldn't make *any* sense for a company continually earning money and paying dividends. So indeed, "sometime" *before* a given dividend, there should be sort of a built-in "pressure" for a stock to gradually rise in expectation of that next cash dividend... in other words: pressure for the stock to have a potential *Dividend Run*.
And notice we put the word "sometime" in quotes in that last sentence, because there are differing views among different dividend investors about *timeframe* when it comes to capturing Dividend Run effects. Some like to invest (and then also to sell) on specific target dates; others like to employ some form of dollar cost averaging. Some like to invest shortly before ex-div, hold for the dividend, and then sell on or after ex-date (having actually capturing the dividend / received the income). Others like to sell the day *before* ex-date (the last possible day where the buyer of the shares will still be "paying for" the upcoming dividend) with the idea to try and maximize *capital gain*. In this capital-gain-focused scenario, one common timeframe we've seen discussed, is to buy about two weeks (ten trading days) prior to the targeted sale date.
For example, consider the 2.38/share AMGN dividend that went "ex-dividend" on 11/21/25. On the prior trading day — the last day where a seller knows that the buyer of their shares will be expecting that dividend amount — shares of AMGN closed at 336.07. And two weeks (ten trading days) prior to *that*, on 11/06/25, shares closed at a price of 315.59. That means that in the final two-week run-up to the 2.38 dividend, AMGN gained 20.48 in price.
Looking back at the last four dividends paid by AMGN, this strategy would have captured a capital gain in excess of the dividend 3 out of 4 times, with a "Divvy Run" total of +32.4 in capital gains. Incidentally, that *exceeds* the sum total *dividend* amounts across those last four dividends, of 9.52. Here's the data:
| Ex-Dividend | ——Price 2 Weeks Prior—» | ——Price 1 Day Prior—» | Run Gain/Loss | |||
|---|---|---|---|---|---|---|
| 11/21/25 | 2.38 | 11/06/25 | 315.59 | 11/20/25 | 336.07 | +20.48 |
| 08/22/25 | 2.38 | 08/07/25 | 284.70 | 08/21/25 | 295.76 | +11.06 |
| 05/16/25 | 2.38 | 05/01/25 | 283.78 | 05/15/25 | 271.69 | -12.09 |
| 02/14/25 | 2.38 | 01/30/25 | 284.02 | 02/13/25 | 296.97 | +12.95 |
| Div Total: | 9.52 | "Divvy Run" Total: | +32.4 | |||
In about two weeks from now, Amgen Inc (NASD: AMGN) will go ex-dividend for its latest dividend of 2.52/share. Will Dividend Run history repeat itself?
Upcoming Dividend: 2.52/share
Ex-Div Date: 02/13/26
Payment Date: 03/06/26
Dividend Frequency: Quarterly
Full AMGN Dividend History »
As the saying goes, past performance is never a guarantee of future returns. But one thing's for sure: for those investors who count Dividend Runs among the tools in their arsenal, AMGN is a good dividend stock to know about and have on your radar screen with its implied annualized yield of 2.95%.
Stay tuned for future Dividend Run candidates, and if you'd like to receive email alerts right into your inbox, enroll in our free Dividend Alerts feature, courtesy of DividendChannel.com.
##### Also see:
AMPH Price Target
Texas Instruments MACD
Top Ten Hedge Funds Holding DC
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Four leading AI models discuss this article
"AMGN's Dividend Run is an observed but unreliable pattern that does not justify tactical trading around ex-dates."
The article highlights AMGN's historical 'Dividend Run' pattern where the stock rose ahead of three of the last four ex-dates, delivering +32.4 in capital gains versus $9.52 in dividends. With the next ex-date on 02/13/26 at $2.52 (2.95% annualized yield) and current price near $336, it suggests potential for a pre-ex run-up. However, the May 2025 period showed a -12.09 decline, proving inconsistency. AMGN trades at ~15x forward P/E with mid-single-digit EPS growth; the pattern appears more random walk than reliable alpha, especially with biotech sector volatility and patent cliffs on key drugs like Enbrel.
The strongest case against is that the three positive runs are statistically insignificant noise; one large loss in four periods already wipes out the edge once transaction costs and taxes are factored in, and nothing in Amgen's fundamentals guarantees buyers will bid up shares in the next 10-14 days.
"The 'Dividend Run' phenomenon is a spurious correlation that misattributes normal market volatility to dividend mechanics, offering no reliable edge for active traders."
The article attempts to frame a 'Dividend Run' as a predictable alpha-generation strategy, but this is essentially a data-mining exercise. The correlation between a 10-day pre-ex-dividend window and share price appreciation is statistically noisy and ignores broader market beta. Amgen (AMGN) is currently trading at roughly 15x forward earnings, heavily influenced by its obesity drug pipeline (MariTide) and cardiovascular data, not quarterly dividend capture. Relying on a 'run-up' pattern ignores the reality that institutional algorithms price in dividends instantly. Investors focusing on this strategy risk being 'picked off' by market volatility, which can easily exceed the $2.52 dividend value in a single trading session.
If institutional liquidity is thin or market sentiment is euphoric, the psychological 'dividend capture' buying pressure can create a self-fulfilling prophecy that temporarily disconnects the stock from its fundamental valuation.
"The observed 'dividend run' gains are likely indistinguishable from general market momentum during those periods, not a repeatable edge."
This article conflates correlation with causation and ignores survivorship bias. The 'dividend run' thesis rests on a 4-sample dataset where 3 of 4 showed gains—statistically meaningless. More critically: AMGN gained 20.48 in the two weeks before ex-div on 11/21/25, but the broader market (SPY) was up ~2% in that same period. Was this AMGN-specific or just beta? The article never addresses this. The strategy's 32.4 total gain vs. 9.52 in dividends sounds compelling until you realize it's measuring price appreciation during a period when biotech rallied broadly. Finally, the 2.95% yield is modest—transaction costs and tax drag could easily eliminate any edge, especially for short-term trading.
If this pattern were genuinely exploitable, arbitrageurs would have already compressed it away; the fact that it persists in public data suggests either it's noise, or the real money already priced it in and retail is chasing ghosts.
"The dividend-run concept for AMGN is not a reliable edge; any pre-ex-date rally is episodic and largely eroded by trading costs and taxes, making it unlikely to produce repeatable alpha."
AMGN's upcoming ex-div date (02/13/26) invites a 'dividend run' thesis, but the pattern rests on fragile logic. Ex-date mechanics imply a price drop roughly equal to the dividend on the ex-date; any pre-ex rally would have to be driven by other factors, and the observed two-week run-ups in a handful of past cycles may be luck, not a repeatable edge. This is a small-sample, post-hoc look that ignores market-wide moves, sector catalysts, and dividend taxes/trading costs. In short, even if AMGN nudges higher ahead of the ex-date, the net, risk-adjusted payoff is unlikely to justify a strategy without wider validation across cycles and costs.
That pre-ex-date rally could be noise or a result of other catalysts; after costs and taxes, the net alpha from chasing a 2-week window is likely negative or near zero.
"Patent erosion in 2026-27 will dominate any short-term pre-ex dividend noise."
Claude correctly flags beta contamination but underweights the patent-cliff asymmetry: Enbrel and Otezla erosion accelerates in 2026-27, compressing the very multiple that makes the 15x P/E look reasonable. A 2-week 'run' cannot offset the 18-24 month derating risk that arrives post-patent expiry regardless of dividend mechanics.
"The market is mispricing Amgen by focusing on legacy patent cliffs while ignoring the binary upside potential of the MariTide pipeline."
Grok, your focus on patent cliffs is vital, but you're missing the 'MariTide' optionality. While Enbrel's erosion is baked into the 15x multiple, the market is currently pricing Amgen as a legacy slow-grower, ignoring the potential for a massive valuation re-rating if obesity data surprises. The 'dividend run' is indeed noise, but the real risk here isn't just patent decay—it's the binary outcome of clinical trials that makes short-term dividend chasing look like rearranging deck chairs on a high-stakes biotech play.
"Optionality on MariTide and patent-cliff risk are separate valuation drivers; neither justifies chasing a 2.52 dividend in a 14-day window."
Gemini's MariTide pivot is clever but sidesteps the core problem: obesity drug optionality doesn't rescue a dividend-capture strategy—it's orthogonal. The real issue nobody's addressed: if MariTide data is truly binary and high-stakes, AMGN's volatility spikes into ex-div windows, *increasing* downside risk for dividend chasers. You're conflating two separate theses. Patent cliffs compress multiples; MariTide could re-rate them. Neither validates a 10-day trading window.
"Arbitrage erosion of a dividend-run edge ignores microstructure frictions; the resulting short-term upside is not a robust edge given patent cliffs and MariTide volatility."
Claude correctly flags beta and survivorship bias, but the stubborn claim that arbitrage would erase a dividend-run edge ignores market microstructure frictions and liquidity gaps around ex-dividends. If AMGN's window coincides with data readouts or fund rebalancing, small persistent edges can survive retail flows. Still, the real risk is the 18-24 month patent cliff and MariTide volatility—short-term upside may pay for large downside risk.
The panel consensus is that relying on a 'dividend run' strategy for Amgen (AMGN) is unreliable and risky, given the small sample size, market beta contamination, and the looming patent cliffs on key drugs. The upcoming ex-dividend date (02/13/26) may not result in a significant pre-ex run-up, and even if it does, the net, risk-adjusted payoff is unlikely to justify the strategy.
The potential re-rating of Amgen's valuation if obesity drug MariTide's clinical trials yield positive surprises.
The 18-24 month patent cliff on key drugs like Enbrel and Otezla, which could compress the company's valuation and offset any short-term gains from a dividend run.