FORTUNE — Smartphone sales, like presidential debates, are a game of expectations — although in the case of Apple’s (AAPL) iPhone 5, it’s one the analysts played but the company didn’t.
In its Monday morning press release, Apple merely announced that it sold a record 5 million iPhone 5s in the three days after Friday morning’s launch, selling out its initial supply. That compares with 4 million iPhone 4S units sold in the same time period last year.
“Demand for iPhone 5 has been incredible and we are working hard to get an iPhone 5 into the hands of every customer who wants one as quickly as possible,” said Tim Cook. “We appreciate everyone’s patience.”
The market was not so patient. The stock opened at $686.91, down $13.18 (1.9%) from Friday’s close, and the analysts who had predicted sales as high as 10 million units had some explaining to do.
Below: Excerpts from the notes to clients we’ve seen so far. As new ones come in, we’ll put them on top.
Wells Fargo’s Maynard Um: Good, Not Great? No, It Could Still Be Great. “We would note that (1) Apple will have 20 more country launches this Friday (9/28), where, based on our interpretation, means there have been no iPhone 5s sold yet in those countries (though Apple should recognize these as revenue because of its sell-in recognition policy), (2) our expectation was lower and also incorporated lower iPhone 4S sales due to the 5 launch (hence, we still believe there is upside potential to our estimates), and (3) while it may make sense to assume units below 6 million units sold over the first three days (the interpretation of “over 5 million” sold), Apple has, in the past, also been conservative in its interpretation.”
Barclay’s Ben Reitzes: iPhone 5 Seeing Significant Constraints, As We Expected. “We believe the key issue in today’s shipment data is in-cell supply constraints. Today’s announcement fits with our view that Apple would only be able to ship limited quantities in the September quarter; we expect a solid production ramp and improved availability in C4Q12. Our Asia team believes Apple has improved the sensitivity of its in-cell screens via a firmware update. As a result the visibility for iPhone 5 has improved and we believe there is less downside risk to production numbers going forward.”
BMO Capital Market’s Keith Bachman: Our view. “We believe iPhone 5 supply is the issue, not demand. Consistent with past launches, most of the iPhone sales are being fulfilled by Apple directly, including online, instead of retail partners. iPhones sold directly by Apple are recognized on a ship-in basis, or only upon customer receipt of the product, while iPhones sold to retail partners are recognized on a ship-out basis. Hence, the actual number of selling days of the iPhone 5 in the September Q may be fewer than the nine days, if most of the phones require two to three days of shipping.”
Merrill Lynch’s Scott Craig: Positive view on iPhone 5 cycle intact; Maintain Buy. “Apple shares are underperforming [Monday] (down 1%+ vs. the S&P 500’s -0.3%), as the 5mn+ iPhone 5 units sold over the past weekend were below the Street’s heightened expectations. We believe most were expecting 6-10mn units over the first weekend, which, in our view, at the high end, did not fully take into consideration the limited initial supply. We believe overall demand remains strong and our estimates reflect most of these supply constraints, leading us to maintain our Buy rating and our above-consensus estimates.”
Topeka Capital’s Brian White (Tuesday follow-up): Don’t Be Fooled by 3-Day iPhone 5 Tally, This is an Epic Launch. “In light of yesterday’s softer than expected iPhone 5 sales over the first three days in retail stores, we thought it would be helpful to put this launch into perspective versus the iPhone 4S on a few other metrics. We believe indicators such as 24-hour pre-orders for the iPhone 5 vs. the 4S, along with pre-order sell out time, post pre-order shipping time and time to sell out at retail stores are more indicative of the strength of the iPhone 5 cycle than yesterday’s three-day sales number that we believe was held back by supply constraints.” See here.
Stifel and Nicholaus’s Aaron C. Rakers: iPhone 5 Launches To Massive Crowds and Positive Reviews; Raising Estimates. “Apple is clearly benefiting from a quicker international launch for iPhone 5 (nine launch countries and 22 more countries in one week) versus 4S (seven launch countries and 22 more countries in two weeks), as well as media attention which has fueled already-robust enthusiasm for iPhone 5. Our analysis, however, suggests that store stock-outs and confirmed shipment delays of 3-4 weeks for online orders will limit unit shipments for C3Q12 to our 21.7mn estimate. That said, with broadening evidence of favorable customer demand we are raising our C4Q12 iPhone unit estimate from 47.4mn to 48.9mn (in-line with consensus) and our CY13 unit estimate from 171.7mn to 172.9mn.”
ISI Group’s Brian Marshall: Clarification to Quick Thoughts on AAPL’s Opening Weekend iPhone 5 Sales. “We expected 6-8 million over the opening weekend… Clarification – recall the 5mil+ iPhone 5 reported sales only takes into consideration: 1) what was sold into partners (e.g., retail outlets, carriers, etc.), 2) sold in AAPL retail stores, and 3) direct to customers only if they signed for the device. Importantly, this doesn’t take into consideration units in delivery direct to customers (i.e., AAPL must have signature of acceptance by customer before it is counted as a sale) and we estimate units in transit could be in the millions currently. We believe the negative pre-market action in AAPL (i.e., down ~2%) is an over-reaction and would be buyers of the stock here.”
Piper Jaffray’s Gene Munster: Launch Weekend Figures Impeded by Availability. “This morning Apple announced the sale of 5 million iPhone 5s in the first weekend compared to our 8 million estimate (6-10 million range). We believe there are two factors that negatively impacted the number. First, our sales expectation assumed that Apple would include all phones pre-ordered online. We believe that this may have been up to 1 million additional units as units pre-ordered after the middle of the first day were projected to be available in October. Second, we noted 1.25 days of Apple Retail inventory compared to 2.5 days during the 4S launch. Our 8 million estimate assumed full weekend availability and the counting of all online pre-orders. We believe that if supply were not a constraint and Apple included all pre-orders, the launch weekend number would have been closer to 7-8 million, assuming ~1 million October pre-order sales and an additional 1-2 million units at retail.”
Sterne Agee’s Shaw Wu: Classic Case of Unrealistically High Expectations. “We are not overly concerned with this ‘disappointing’ number as we believe this is a classic case of near-term expectations getting out of touch with reality. We find it unfortunate that some analysts continue to publish irresponsible estimates without taking into account realistic demand trends and potential supply constraints on new in-cell touchscreens.”
Topeka Capital’s Brian White: Three Day iPhone 5 Sales Top 5 Million but Supply Holds Back More Sales. “This morning, Apple announced that iPhone 5 sales exceeded 5 million during the first three days of retail availability and below our expectations (we were at the lower end of expectations). We believe this shortfall is largely due to supply availability and the fact that most consumers are opting for the pre-order option. We remain aggressive buyers of Apple on any weakness in the stock price this morning as we believe Apple has another blockbuster on its hands with the iPhone 5 and we expect the “iPad Mini” to launch in the coming weeks.”
Asymco’s Horace Dediu, who had predicted first weekend sales of 6 million, offered five points by e-mail:
1. They were only able to offer 25% more units over the launch weekend than what they could offer with the 4S. That implies that either they are not fully ramped or their ramp started later than last year. This could be because it’s a new product vs. last year’s modest changes from the 4 or it could be that they are constrained on components.
2. In retrospect it seems they knew they would be constrained in the opening because they did not expand the footprint from last year. The country count went up by 2 but did not add big countries.
3. The demand is far higher than last year given the higher number of pre-orders and anecdotal counts of people waiting in line.
4. The capex story implies that production should be at double the rate from last year, but that does not help with launch. If they produce twice as much but have half the lead time then the product on hand is the same as last year. I suspect the ramp is a bit slower than they had hoped. The press release seems apologetic about them not being able to meet demand.
5. They are more aggressive in terms of country launches by end of year so maybe they will make up for the shortfalls during the fourth.