Sam Alich, Plaintiff, v. Opendoor Technologies Incorporated, et al., Defendants.
No. CV-22-01717-PHX-MTL
IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF ARIZONA
February 28, 2024
WO
Opendoor is a publicly traded company that uses advanced technology to buy and resell homes. Plaintiffs are purchasers of Opendoor stock. They are suing Opendoor, Opendoor personnel, and the Underwriters of Opendoor‘s secondary public offering for securities fraud.
This matter is before the Court on the Opendoor and Underwriter Defendants’ Motions to Dismiss the Consolidated Amended Complaint.1 (Doc. 48 and 51.) Additionally, the Opendoor Defendants request that the Court consider documents referenced in the Consolidated Amended Complaint in its analysis of the Motions. (Docs. 49, 50.) The Motions are fully briefed. (Docs. 48, 51, 55, 56, 60, 62.) The Court held oral argument on January 17, 2024. For the following reasons, the Court considers
the referenced documents under the doctrine of incorporation by reference and dismisses
I. FACTS
In the past decade, rapid technological development has shaped and reshaped commerce. Artificial intelligence is now disrupting industries throughout the world economy. The ways that people shop, consume entertainment, move about, and vacation have fundamentally changed. So, too, has the way that people conduct major transactions such as home selling and buying.
iBuying, short for instant buying, is a relatively recent phenomenon intended to simplify major transactions. (Doc. 39 ¶¶ 51, 52.) Powered by advanced technology, iBuyers, or businesses engaged in iBuying, use automated valuation models to price homes. (Id. ¶ 51.) Based on those valuations, iBuyers attempt to quickly purchase and resell homes. (Id.)
Founded in 2014, Opendoor is an iBuying company operating in the residential real estate market. (Id. ¶¶ 51-52.) Its process follows the typical structure of an iBuying business. Individuals looking to sell their homes can go to Opendoor‘s website or mobile app, input some relevant data points, and receive an “instant” cash offer generated by Opendoor‘s proprietary algorithm. (Id. ¶ 53.) After making the initial offer, Opendoor personnel conduct a virtual assessment and adjust the offer based on the home‘s condition. (Id.) Then, Opendoor makes a final offer, which is reduced by a 5% service charge. (Id.) If the homeowner accepts, Opendoor pays the seller, takes ownership of the home, makes repairs if necessary, and lists it for resale. (Id. ¶¶ 55, 75-76.)
Opendoor‘s algorithm is central to the success of its iBuying business. (Id. ¶¶ 56, 59.) To consistently sell homes at a profit, Opendoor‘s algorithm must produce generally accurate initial offers. (Id. ¶ 4.) While those initial offers will often be adjusted after the homes’ unique defects are revealed through an inspection, they must provide Opendoor with a viable starting point. (Id. ¶¶ 4, 53.) Accordingly, the algorithm must account for the value of the homes at the time of the initial offer and accurately forecast their future value. (Id. ¶¶ 4, 7, 60.)
Opendoor has worked toward this objective since its inception. (Id. ¶¶ 3, 52, 56.) It began as a private company, but in 2020, went public via a reverse merger with Defendant Social Capital Hedosophia Holdings Corp. II, a Special Purpose Acquisition Company (“SPAC“).2 (Id. ¶ 324.) After the de-SPAC merger, the surviving entity changed its name to Opendoor Technologies Inc. (Id. ¶ 325.) Then, in December of that year, Opendoor published offering documents which registered the first issuance of shares of Opendoor common stock. (Id.) On December 21, 2020, the newly issued Opendoor shares began publicly trading on the NASDAQ stock exchange. (Id.) In February 2021, Opendoor published new offering documents registering the issuance of additional stock. (Id. ¶ 327.) On February 9, 2021, those shares began publicly trading. (Id.)
As Opendoor informed stockholders, its iBuying business is its main revenue driver. (Id. ¶ 55.) In evaluating the financial health of that business, Opendoor told its investors that “[t]he ultimate measure [they] should hold [Opendoor] accountable for is how [it is] doing on contribution margin delivery.” (Id. ¶ 8.) Contribution margin refers to a metric used “to understand how a specific product contributes to [a] company‘s profit.” Amy Gallo, Contribution Margin: What It Is, How to Calculate It, and Why You Need It, Harvard Business Review (Oct. 13, 2017), https://hbr.org/2017/10/contribution-margin-what-it-is-how-to-calculate-it-and-why-you-need-it. In this case, the “product” is Opendoor‘s iBuying operation.
Opendoor‘s contribution margin was strong immediately following its public
Lead Plaintiffs are retirement service providers that purchased Opendoor stock on behalf of their members and beneficiaries during the Class Period.3 (Id. ¶¶ 356-57.) Stuart Graham Hereford, another Plaintiff, purchased stock in Social Capital Hedosophia Holdings Corp. II and, after the de-SPAC merger, from Opendoor on several occasions throughout the Class Period. (Doc. 39-3.) Together, Plaintiffs allege that Defendants made several materially misleading statements during the Class Period, that those statements artificially inflated Opendoor‘s stock price, that they relied on those statements, and that they suffered damages when the stock price dropped after the truth was revealed. (See generally Doc. 39.)
Against the Opendoor Defendants, Plaintiffs assert claims under
II. LEGAL STANDARD
To survive a motion to dismiss pursuant to
Securities fraud suits face heightened pleading standards. “At the pleading stage, a complaint stating claims under [S]ection 10(b) and Rule 10b-5 must satisfy the dual pleading requirements of
Because allegations of fraud inescapably carry a degree of moral turpitude,
Rule 9(b) imparts a heightened note of seriousness, requiring a greater degree of pre-discovery investigation by the plaintiff, followed by the plaintiff‘s required particular allegations, thereby protecting a defendant‘s reputation from frivolous and unfounded allegations and permitting a particularized basis for a defendant to respond to the particularized allegations.
Irving Firemen‘s Relief & Ret. Fund v. Uber Techs., Inc., 998 F.3d 397, 404 (9th Cir. 2021) (citation omitted).
Similarly, the PSLRA requires that “the complaint shall specify each statement alleged to have been misleading, the reason or reasons why the statement is misleading, and, if an allegation regarding the statement or omission is made on information and belief, the complaint shall state with particularity all facts on which that belief is formed.”
III. DISCUSSION
A. Section 10(b)
“Section 10(b) of the Exchange Act prohibits manipulative or deceptive practices in connection with the purchase or sale of a security.” In re Facebook, Inc. Sec. Litig., 87 F.4th 934, 947 (9th Cir. 2023) (internal marks omitted).
To state a claim under
1. Actionable False or Misleading Statements
The Opendoor Defendants first argue that Plaintiffs’ Section 10(b) claim fails because Plaintiffs have not alleged that they made any actionable false or misleading statements. (Doc. 48 at 18-28; Doc. 60 at 7-18.) In response, Plaintiffs point to several categories of statements made by the Opendoor Defendants alleged in the Consolidated Amended Complaint to have been false or misleading. (Doc. 56 at 15-21.)
“A plaintiff can allege falsity by ‘point[ing] to [the] defendant‘s statements that directly contradict what the defendant knew at the time.‘” Jaeger v. Zillow Grp., Inc., 644 F. Supp. 3d 857, 870 (W.D. Wash. 2022) (quoting Khoja v. Orexigen Therapeutics, Inc., 899 F.3d 988, 1008 (9th Cir. 2018)). “[A] statement is misleading if it would give a reasonable investor the ‘impression of a state of affairs that differs in a material way from the one that actually exists.‘” Berson v. Applied Signal Tech., Inc., 527 F.3d 982, 985 (9th Cir. 2008) (quoting Brody v. Transitional Hosps. Corp., 280 F.3d 997, 1006 (9th Cir. 2002)). “To be misleading, a statement must be ‘capable of objective verification.‘” Retail Wholesale & Dep‘t Store Union Local 338 Ret. Fund v. Hewlett-Packard Co., 845 F.3d 1268, 1275 (9th Cir. 2017) (quoting Or. Pub. Emps. Ret. Fund, 774 F.3d at 606).
i. Statements Regarding the Pricing Algorithm
Plaintiffs allege that the Opendoor Defendants, through their challenged statements, “created the impression that [Opendoor‘s] success was the result of its AI-powered algorithm, which they claimed drove pricing decisions and kept Opendoor profitable in any housing market by reacting to changing economic conditions.” (Doc. 55 at 15.) Plaintiffs challenge the following statements regarding the accuracy and operability of Opendoor‘s pricing algorithm:
- The algorithm uses “machine learning to drive pricing decisions.” (Doc. 39 ¶ 169.)
- Opendoor‘s “proprietary, machine learning-based pricing models are key to [Opendoor‘s] ability to acquire and resell thousands of homes per month accurately, profitably, and with increasing levels of automation.” (Id. ¶ 173.)
- The algorithm can “dynamically adjust to leading market indicators and react to real-time macro- and micro-economic conditions.” (Id. ¶¶ 169, 177.)
- The algorithm‘s ability to “dynamically price homes in response to both micro and macro factors [] allows [Opendoor] to manage . . . within [a] 4% to 6%, annual [contribution] margin range, [] regardless of the home cycle [Opendoor is] operating in.” (Id. ¶ 186.)
(Doc. 55 at 15.)
Plaintiffs allege that these statements were misleading because “the algorithm could not accurately price homes (even in a hot market) and as such [Opendoor‘s] pricing
The Opendoor Defendants counter that Plaintiffs’ allegations conflate their business practices in the initial and final offer stages. (Doc. 48 at 20-21 n.11.) They argue that Opendoor never claimed that all its pricing decisions were driven by its algorithm. (Id. at 20-21.) Instead, Opendoor‘s initial offers were driven primarily (though not entirely) by its algorithm, but creating final offers required the aid of substantial human involvement. This makes sense. While the pricing algorithm took the lead in generating an initial offer based on available data, that initial offer had to be adjusted by the unique characteristics that each individual property possessed—such as damaged or aged aspects that may have necessitated repairs. This latter task naturally required personal intervention.
The Opendoor Defendants say that they disclosed the existence of this human involvement in the following statements:4
- “The majority of our initial offers are algorithmically generated and do not require any human intervention.” (Doc. 49-1 at 45.)
“After all the data has been collected and incorporated, each offer is reviewed and finalized by members of our pricing team, allowing us to marry the best of algorithmic insights with human judgment.” (Doc. 49-2 at 10.) - “We appraise and price homes we buy and sell using data science, proprietary algorithms, and analysis from specially trained employees.” (Id. at 19.)
- “I think there‘s a sense that it‘s just about the model . . . . But we‘ve also done a really good job of integrating systematic modeling with [] the human insight and the human overlay.” (Doc. 49-3 at 10.)
(Doc. 48 at 20-21.)
The Opendoor Defendants also argue that Plaintiffs’ confidential witnesses corroborate the disclosures. (Doc. 60 at 10.) For example, CW 2‘s statement that about 90% of final offers were generated by pricing analysts is consistent with Opendoor‘s disclosure that its final offers come after human adjustments. CW 5‘s estimate that about 50% of initial offers were reviewed by humans is generally consistent with Opendoor‘s disclosure that most, though not all, of its initial offers do not involve human intervention.
Plaintiffs, however, contend that the Opendoor Defendants’ disclosures fall short because they do not “disclose[] that Opendoor was forced to rely on humans to price its offers because its algorithm could not accurately price homes . . . [or] that [Opendoor] was engaged in deceptive practices . . . that contributed to [Opendoor‘s] profitability.” (Doc. 55 at 17.) Additionally, Plaintiffs argue that the “[t]he instant case is strikingly analogous to the securities fraud lawsuit against Zillow [in Jaeger, 644 F. Supp. 3d 857], which investors brought after Zillow left the iBuying business in November 2021.” (Id.) In both cases, say Plaintiffs, the companies relied on extensive human involvement in their pricing scheme while misleading investors into believing that their algorithms drove their success. (Id. at 18.)
Plaintiffs’ arguments are unavailing. First, Jaeger is distinguishable. There, Zillow sought to rapidly scale up its iBuying operation in an attempt to catch up to its competitors. Jaeger, 644 F. Supp. 3d at 865-66. To do so, Zillow devised a plan that it called “Project
Plaintiffs’ attempt to frame Opendoor‘s conduct in a similar light is unconvincing. They allege that, like Zillow, Opendoor concealed its practices of adjusting initial offers in the final offer stage and failing to refund deductions made for anticipated repairs that became unnecessary. (Doc. 39 ¶¶ 170-74; Doc. 55 at 18.) But unlike Project Ketchup, those practices were sufficiently disclosed to investors and consistent with the Opendoor Defendants’ public statements. (Doc. 49-1 at 10, 45; Doc. 49-2 at 10, 11, 19; Doc. 49-3 at 10.)
The Court also finds that, even taking the facts in the light most favorable to Plaintiffs, the four challenged statements are not false and would not have misled a reasonable investor. The Opendoor Defendants’ statement that the algorithm used “machine learning to drive pricing decisions” is a simple statement about Opendoor‘s use of AI in its pricing scheme and does not suggest that Opendoor personnel lack any significant role in that process. Nor does the Opendoor Defendants’ representation that Opendoor‘s algorithm is “key” to the success of its business model. The other two challenged statements—about the algorithm‘s ability to dynamically adjust and keep Opendoor within a 4% to 6% annual contribution margin—likewise would not lead reasonable investors to conclude that Opendoor‘s entire process, from initial offer to final offer, was driven by its algorithm. This is especially true in light of the Opendoor Defendants’ disclosures. As discussed below, however, Plaintiffs successfully pleaded these two latter statements as misleading insofar as they suggest that Opendoor‘s algorithm
ii. Statements Regarding Profitability
Plaintiffs next argue that the Opendoor Defendants misled investors by guaranteeing Opendoor‘s “ability to remain profitable in any housing market.” (Doc. 55 at 19.) Plaintiffs challenge the following statements:
- The algorithm can “dynamically adjust to leading market indicators and react to real-time macro- and micro-economic conditions.” (Doc. 39 ¶¶ 169, 177.)
- “[O]ur model really works in upmarkets, it‘s going to work in flat markets, it‘s going to work in downmarkets.” (Id. ¶ 181 (spelling in original).)
- The algorithm‘s ability to “dynamically price homes in response to both micro and macro factors [] allows [Opendoor] to manage . . . within [a] 4% to 6%, annual [contribution] margin range, [] regardless of the home cycle [Opendoor is] operating in.” (Id. ¶ 186.)
- “[I]f you run the business model through the worst recession in U.S. history, we would still have positive contribution margin.” (Id. ¶ 189.)
- Opendoor “ha[s] good evidence that [it] can operate in [a] declining market.” (Id. ¶ 192.)
- Opendoor‘s “systems are doing exactly what they‘re designed to do, which is responding very quickly, adjusting prices to market.” (Id. ¶ 199.)
(Doc. 55 at 15, 19-20, 23.)
According to Plaintiffs, “the algorithm never worked as Defendants claimed.” (Doc. 55 at 20.) And, according to Plaintiffs, many of the same confidential witness statements cited above support this contention. (Supra at 8; see also Doc. 39 ¶¶ 62-73; Doc. 55 at 22.)
The Opendoor Defendants argue that the challenged statements are not misleading because their many disclaimers notified reasonable investors that the algorithm could not guarantee profitability. (Doc. 48 at 19-20.) They cite the following statements from the Opendoor Defendants as disclosing the risk that Opendoor could incur losses:
- “Our business is dependent upon our ability to accurately price and portfolio
manage inventory and an ineffective pricing or portfolio management strategy may have a material adverse effect on our business.” (Doc. 49-1 at 10-11.) - Opendoor could “incur significant losses in the future” due to its “failure to appropriately price and manage the home inventory [it] acquire[d].” (Doc. 49-2 at 17.)
- Opendoor “may be unable to liquidate . . . inventory at prices that allow [it] to meet [its] margin targets or recover [its] costs.” (Doc. 49-1 at 11.)
- Opendoor‘s “success depends, directly and indirectly, on general economic conditions, [and] the health of the U.S. residential real estate industry, particularly the single family home resale market.” (Id. at 7-8.)
- “Home prices can be volatile and the values of [Opendoor‘s] inventory may fluctuate significantly.” (Id. at 14.)
- “Mortgage rates are currently low as compared to most historical periods,” and rate increases “could result in a decline in the demand for [Opendoor‘s] homes” and “adversely affect [Opendoor‘s] business or financial results.” (Id. at 15-16.)
- “[C]hanging market conditions are reflected in [Opendoor‘s] pricing for new acquisitions, largely leaving previously acquired inventory at risk to potential market volatility.” (Doc. 49-2 at 12.)
(Doc. 48 at 13, 19-20.)
Plaintiffs argue that the Opendoor Defendants’ disclosures were insufficient because (1) they did not correct the impression that the algorithm “worked in all housing markets;” and (2) they did not disclose that the algorithm already failed to accurately estimate home value. (Doc. 55 at 21-22.) The first argument can be quickly dismissed. It stands to reason that no reasonable investor would ever believe that a particular business model, even one premised on impressive technology, is guaranteed to be profitable in all circumstances.
Plaintiffs also rely on statements from financial analysts writing for Wedbush, InvestorPlace, and Seeking Alpha to argue that reasonable investors believed Opendoor‘s
The statements do, however, reflect an impression that Opendoor‘s algorithm was, at the very least, generating roughly accurate initial offers at the time the statements were made. If the algorithm was in fact already failing to accurately estimate home values, the analysts cited in the Consolidated Amended Complaint were incorrect in their impressions. Plaintiffs adequately allege that the analysts’ erroneous conclusions were due to the Opendoor Defendants’ purportedly misleading statements regarding the efficacy of their algorithm. (Id. ¶¶ 96-99, 112.) Because “[t]he perceptions of analysts are an acceptable measure of what reasonable investors would have understood,” Jaeger, 644 F. Supp. 3d. at 872, the Court finds that, taking Plaintiffs’ allegations as true, a reasonable investor would have been misled by the Opendoor Defendants’ statements into believing that Opendoor‘s algorithm was accurately estimating and forecasting home values at the time the statements were made.
The Opendoor Defendants argue that their misleading statements are nonetheless non-actionable puffery. (Doc. 48 at 24-27.) “In the Ninth Circuit, vague, generalized assertions of corporate optimism or statements of mere puffing are not actionable material misrepresentations under federal securities laws because no reasonable investor would rely on such statements.” In re Restoration Robotics, Inc. Sec. Litig., 417 F. Supp. 3d 1242, 1255 (N.D. Cal. 2019) (internal quotations and citations omitted). “When valuing corporations, . . . investors do not rely on vague statements of optimism like ‘good,’ ‘well-regarded,’ or other feel good monikers.” In re Cutera Sec. Litig., 610 F.3d 1103,
The Court finds that the Opendoor Defendants’ statements are too concrete and verifiable to constitute non-actionable puffery. See e.g., Jaeger, 644 F. Supp. 3d at 872-73. Each of the challenged statements are premised on a concrete, verifiable supposition—Opendoor‘s algorithm, at a bare minimum, can accurately estimate and forecast home values. Opendoor‘s algorithm either does this in up, flat, and down markets or it does not. In the face of rapid economic downturn, it either responds very quickly and adjusts prices to market or it does not. It is either capable of keeping Opendoor within a 4% to 6% contribution margin range or it is not. These elements of concreteness and verifiability were not present in the cases relied upon by the Opendoor Defendants. See Plevy v. Haggerty, 38 F. Supp. 2d 816, 827-28 (C.D. Cal. 1998); Hutton v. McDaniel, 264 F. Supp. 3d 997, 1021 (D. Ariz. 2017); Macomb Cnty. Emps.’ Ret. Sys. v. Align Tech., Inc., 39 F.4th 1092, 1099 (9th Cir. 2022).
Thus, the Court finds that these six challenged statements are actionably misleading because they would have led a reasonable investor to conclude that Opendoor‘s algorithm was accurately pricing homes and adjusting to changes within the market when the statements were made.
iii. Statements Regarding Opendoor‘s Business Model and the 2019 FTC Complaint
Plaintiffs next argue that the Opendoor Defendants misled investors by mischaracterizing its business model. (Doc. 55 at 23-24.) The Opendoor Defendants claimed that Opendoor‘s “business model is designed to generate margins from [Opendoor‘s] service charge to sellers and ancillary products and services associated with
The Opendoor Defendants reply that investors were fully aware that Opendoor generated revenue from reselling homes. (Doc. 48 at 25; Doc. 60 at 17 nn.10, 13-14.) Specifically, Opendoor told investors that it “aim[ed] to maximize resale margin.” (Doc. 49-1 at 45; see also Doc. 49-25 at 54, 62.) Because Opendoor adequately disclosed that it sought to resell homes at a profit, the Court finds that, even taking Plaintiffs’ allegations as true, a reasonable investor would not have been misled by the challenged statement.
Plaintiffs next allege that the Opendoor Defendants misled investors by reassuring them that “the allegations raised by the FTC are related to activity that occurred between 2017 and 2019.” (Doc. 39 ¶ 196.) In fact, say Plaintiffs, the activity continued throughout the Class Period. But Plaintiffs do not allege that the conduct giving rise to the FTC complaint—allegedly misleading advertising practices—continued. (See generally Doc. 39; see also Doc. 49-7.) Instead, they allege, based on the detailed statements of confidential witnesses, that Opendoor personnel continued to manually adjust the algorithm‘s offers and charge sellers for repairs that they did not make. (Id. ¶¶ 62-84, 143-49.) But Opendoor repeatedly disclosed that people had a role in pricing and that repair credits were estimates for which Opendoor bore the risk or reward of staying within budget. (See e.g., Doc. 49-2 at 10, 11.) Therefore, the Court finds that a reasonable investor would not have been misled by this challenged statement.
2. Scienter
To state a claim under
These standards “present no small hurdle for the securities fraud plaintiff.” Schueneman v. Arena Pharms., Inc., 840 F.3d 698, 705 (9th Cir. 2016) (citation omitted); see also Nguyen v. Endologix, Inc., 962 F.3d 405, 414 (9th Cir. 2020) (“The PSLRA‘s strong inference requirement has teeth. It is an exacting pleading obligation.“) (cleaned up). The Court must “engage in a comparative evaluation [and] . . . consider, not only inferences urged by the plaintiff . . . but also competing inferences rationally drawn from the facts alleged.” Tellabs, Inc. v. Makor Issues & Rights Ltd., 551 U.S. 308, 314 (2007). A complaint will survive a motion to dismiss “only if a reasonable person would deem the inference of scienter cogent and at least as compelling as any opposing inference one could draw from the facts alleged.” Id. at 324.
The Opendoor Defendants argue that Plaintiffs have failed to adequately plead scienter. (Doc. 48 at 28-34; Doc. 60 at 18-23.) Plaintiffs disagree. (Doc. 55 at 25-32.) They
i. Defendant Wu‘s Stock Sales
Defendant Wu is a co-founder of Opendoor, the former Chairman of Opendoor‘s Board of Directors, and Opendoor‘s former Chief Executive Officer. (Doc. 39 ¶ 39.) Plaintiffs argue that his sale of a portion of his Opendoor stock is “highly indicative of scienter.” (Doc. 55 at 26; Doc. 39 ¶¶ 239-47.) Setting aside the question of whether a single defendant‘s trading can establish a strong inference of scienter for a group of co-defendants, the Court finds that Defendant Wu‘s sales do not even establish a strong inference of scienter as to him.
“‘To evaluate suspiciousness of stock stales, [the Court] consider[s], inter alia, three factors: (1) the amount and percentage of shares sold; (2) timing of the sales; and (3) consistency with prior trading history.‘” In re Quality Sys., Inc. Sec. Litig., 865 F.3d 1130, 1146 (9th Cir. 2017) (quoting Nursing Home Pension Fund, Loc. 144 v. Oracle Corp., 380 F.3d 1226, 1232 (9th Cir. 2004)).
First, the amount and percentage of shares sold by Defendant Wu do not indicate scienter. Plaintiffs allege that, between November 16 and 18, 2021, Defendant Wu sold 7% of his Opendoor stock. (Doc. 39 ¶ 239; see also Doc. 55 at 29.) They concede that this is a relatively small percentage but argue that “even sales of a relatively small percentage of an insider‘s holdings can support an inference of scienter if the absolute dollar value of the transactions is high.” (Doc. 55 at 29.) For example, Plaintiffs rely upon In re SeeBeyond Technologies Corp. Securities Litigation, 266 F. Supp. 2d 1150 (C.D. Cal. 2003) (hereinafter ”In re SeeBeyond“). But In re SeeBeyond is readily distinguishable, as the court‘s decision relied on the fact that the “defendants . . . admittedly lied to analysts and investors.” 266 F. Supp. 2d at 1169. Not so here. Additionally, the Court finds that courts within the Ninth Circuit routinely find that sales of similarly small percentages of defendants’ stock holdings do not give rise to a strong inference of scienter. See e.g., In re Alteryx, Inc. Sec. Litig., No. 8:20-cv-01540 DOC (JDEx), 2021 WL 4551201, at *4 (C.D. Cal. June 17, 2021); Zamir v. Bridgepoint Educ., Inc., No. 15-CV-408 JLS (DHB), 2016 WL 3971400, at *10 (S.D. Cal. July 25, 2016); In re Juniper Networks, Inc., 158 F. App‘x 899, 901 (9th Cir. 2005).
Nor does the timing of Defendant Wu‘s sales indicate scienter. The parties characterize those sales somewhat differently. As told by Plaintiffs, Defendant Wu‘s sales were suspiciously timed because they were made outside the parameters of his 10b5-1 plan and came just three days after he provided allegedly false reassurances to investors following Zillow‘s exit from the iBuying marketplace.5 (Doc. 55 at 26-27.) But according to the Opendoor Defendants, Defendant Wu had no traditional 10b5-1 plan at all. (Doc. 60 at 20-21.) Instead, his sales were pre-determined and exclusively intended to cover tax liability. (Id.) Zillow‘s exit is irrelevant because that was not due to the failure of its algorithm, but its plan to overprice homes in order to gain market share. (Id. at 21.) Finally, Defendant Wu‘s one-year lockup period under SEC Rule 144 had just expired, leaving him free to unload stock; thus, such rapid sales are not unusual.6 (Id.)
Plaintiffs’ allegations fail to meet the exacting pleading standards imposed by
Neither party addresses the third factor—whether Defendant Wu‘s sales were consistent with his prior trading history. (See Doc. 48 at 28-31; Doc. 55 at 26-29; Doc. 60 at 18-21.) The Consolidated Amended Complaint alleges only that the sales “were by far Defendant Wu‘s largest sales of Opendoor stock historically.” (Doc. 39 ¶ 244.) But it does not provide any “meaningful trading history for purposes of comparison to the stock sales within the Class Period.” Zucco, 552 F.3d at 1005 (cleaned up).
Accordingly, the Court finds that Plaintiffs’ allegations concerning Defendant Wu‘s stock sales fail to support a strong inference of scienter.
ii. Core Operations Doctrine
Plaintiffs next attempt to plead scienter through the “core operations” doctrine. (Doc. 39 ¶¶ 248-53.) This theory “relies on the principle that corporate officers have knowledge of the critical core operation of their companies.” Police Ret. Sys. of St. Louis v. Intuitive Surgical, Inc., 759 F.3d 1051, 1062 (9th Cir. 2014) (internal marks omitted). Core operations may support a strong inference of scienter under three circumstances:
First, the allegations may be used in any form along with other allegations that, when read together, raise an inference of scienter that is cogent and compelling, thus strong in light of other explanations . . . . Second, such allegations may independently satisfy the PSLRA where they are particular and suggest that defendants had actual access to the disputed information . . . . Finally, such allegations may conceivably satisfy the PSLRA standard in a more bare form, without accompanying particularized allegations, in rare circumstances where the nature of the relevant fact is of such prominence that it would be absurd to suggest that management was without knowledge of the matter.
S. Ferry LP, No. 2 v. Killinger, 542 F.3d 776, 785-86 (9th Cir. 2008) (internal marks omitted). “Proof under this theory is not easy.” Intuitive Surgical, Inc., 759 F.3d at 1062. To successfully allege scienter under the core operations doctrine, “[a] plaintiff must produce either specific admissions by one or more corporate executives of detailed
Plaintiffs’ arguments regarding the core operations doctrine are brief and ambiguous. (See Doc. 39 ¶¶ 248-53.) For example, Plaintiffs state that “it is implausible to suggest that Defendants did not know about the human-driven process for pricing.” (Doc. 55 at 30 (cleaned up).) But they also state, much more broadly, that “it would be absurd to suggest that management was without knowledge of the algorithm‘s true capabilities.” (Id. (quoting Reese v. Malone, 747 F.3d 557, 577 (9th Cir. 2014).) In either case, the Consolidated Amended Complaint lacks the necessary particularized allegations to plead scienter through the core operations doctrine.
While the Consolidated Amended Complaint adequately alleges that the algorithm was the core of Opendoor‘s operations, Plaintiffs have not “produce[d] either specific admissions by one or more corporate executives of detailed involvement in the minutia of [Opendoor‘s pricing] operations . . . or witness accounts demonstrating that executives had actual involvement in creating false reports.” Intuitive Surgical, Inc., 759 F.3d at 1062. In the Consolidated Amended Complaint, Plaintiffs generally assert that because the algorithm was so crucial to the success of the business, the Opendoor Defendants must have been aware of the algorithm‘s efficacy. (See Doc. 39 ¶¶ 248-53.) In their response brief, Plaintiffs additionally argue that the 2019 FTC investigation into Opendoor must have made the Opendoor Defendants aware of the algorithm‘s present inability to estimate and forecast home values accurately. (Doc. 55 at 30.)
The Court finds that Plaintiffs’ allegations are insufficient to establish a strong inference of scienter through the core operations doctrine. First, their vague allegations based on the algorithm‘s general importance are insufficient. See Palm Harbor Special Fire Control & Rescue Dist. Firefighters Pension Plan v. First Solar Inc., No. CV-22-00036-PHX-MTL, 2023 WL 4161355, at *5 (D. Ariz. June 23, 2023) (“Although Plaintiffs argue that it ‘would be absurd to suggest that Defendants were not aware of the problems with the Series 6 module[,]’ the law requires more in the form of specific admissions or”
witness accounts.“). Similarly, Plaintiffs’ reliance on the FTC investigation is misplaced. The FTC investigation had nothing to do with the accuracy of Opendoor‘s algorithm. (See generally Doc. 49-7.) Thus, the core operations doctrine is no help to Plaintiffs.iii. Zillow‘s Exit from the iBuying Marketplace
Plaintiffs also argue that Zillow‘s exit from the iBuying marketplace supports an inference of scienter. (Doc. 39 ¶¶ 273-75; Doc. 55 at 30-31.) They say that because Zillow was a major competitor, presumably with expansive resources, but nonetheless exited the iBuying space “because of its failure to accurately price homes” (Doc. 39 ¶ 273), the Opendoor Defendants “knew, or were reckless in not knowing, that their algorithm was not able to accurately price homes in any environment” (id. ¶ 274).
The Court is unpersuaded. Plaintiffs merely insinuate that, following Zillow‘s exit, the Opendoor Defendants either did or should have concluded that if Zillow cannot succeed, they cannot either. But Plaintiffs’ allegations fail to specify with the level of particularity demanded by the
iv. FTC Investigation
Next, Plaintiffs allege that the FTC investigation into Opendoor‘s operations supports an inference of scienter. (Doc. 39 ¶¶ 264-72; Doc. 55 at 31.) According to Plaintiffs, because the FTC investigated “the very conduct that the [Consolidated Amended] Complaint alleges was withheld from investors” and because “Opendoor had conducted internal analyses showing its employees manually adjusted offers and charged for unnecessary repairs,” the Opendoor Defendants must have been aware of the undisclosed practices. (Doc. 55 at 31.) While the FTC investigation dealt with alleged activities before the Class Period, Plaintiffs allege that those activities continued through the Class Period. (See e.g., Doc. 39 ¶¶ 62-84, 143-49.)
v. The Opendoor Defendants’ Statements
Plaintiffs finally argue that the Opendoor Defendants’ statements regarding the algorithm‘s capabilities, references to internal testing of the algorithm, and timing of their statements support an inference of scienter. (Doc. 39 ¶¶ 108, 181, 254-63; Doc. 55 at 31.) But again, the Consolidated Amended Complaint lacks particularized allegations that the Opendoor Defendants possessed contradictory information about the efficacy of the algorithm. (See generally Doc. 39 ¶¶ 254-63.) For example, Plaintiffs do not specifically allege that Opendoor‘s internal testing revealed that the algorithm was not accurately estimating or forecasting home values. (See generally id.) Nor do they allege that such troubling test results, if indeed they existed, were conveyed to the Opendoor Defendants. (See generally id.) Thus, Plaintiffs cannot establish a strong inference of scienter on this theory either.
Accordingly, the Court finds that Plaintiffs’ allegations do not meet the demanding requirements imposed by
3. Loss Causation
Even if Plaintiffs had adequately pleaded a strong inference of scienter, their
“Typically, to establish loss causation, a plaintiff must show that the defendants’ alleged misstatements artificially inflated the price of stock and that, once the market learned of the deception, the value of the stock declined.” Irving Firemen‘s Relief & Ret. Fund, 998 F.3d at 407 (citing Nuveen Mun. High Income Opportunity Fund v. City of Alameda, Cal., 730 F.3d 1111, 1119-20 (9th Cir. 2013)). This is known as the “fraud-on-the-market” theory. Nuveen, 730 F.3d at 1120. To advance this theory, “the plaintiff must show that after purchasing her shares and before selling, the following occurred: (1) the truth became known, and (2) the revelation caused the fraud-induced inflation in the stock‘s price to be reduced or eliminated.” In re BofI Holding, 977 F.3d at 789 (internal marks omitted). “The most common way for plaintiffs to prove that ‘the truth became known’ is to identify one or more corrective disclosures.” Id. at 790 (citing Mineworkers’ Pension Scheme v. First Solar Inc., 881 F.3d 750, 753-54 (9th Cir. 2018) (per curiam)). Corrective disclosures occur “when ‘information correcting the misstatement or omission that is the basis for the action is disseminated to the market.‘” Id. (quoting
While determining what constitutes a corrective disclosure “has proved more challenging than might have been expected, a few basic ground rules can be sketched out.” Id. at 790. Corrective disclosures do not require “an admission of fraud by the defendant or a formal finding of fraud by a government agency.” Id. (citing Metzler, 540 F.3d at 1064). Rather, a corrective disclosure can “come from any source, including knowledgeable third parties such as whistleblowers, analysts, or investigative reporters.”
Plaintiffs allege that three corrective disclosures occurred and that significant losses ensued for shareholders. (Doc. 39 ¶¶ 210-236.)
i. Decreasing Contribution Margin
On February 24, 2022, Opendoor announced its fourth quarter and full year 2021 financial results. (Id. ¶ 220.) Opendoor revealed that its fourth quarter 2021 contribution margin was 4%, representing a significant decrease from its 12.6% contribution margin in the fourth quarter of 2020. (Id. ¶ 221.) On February 25, 2022, Opendoor‘s stock price closed at $8.44, representing a $2.54 loss from the previous trading day. (Id. ¶ 222.) Plaintiffs allege that Opendoor‘s announcement “started to reveal that Defendants had overstated the purported benefits and competitive advantages of the algorithm and that [Opendoor] was equally susceptible to the same market fluctuations as every other iBuying and real estate [c]ompany in the world.” (Id. ¶ 220.)
The Court finds that the February announcement of Opendoor‘s contribution margin did not constitute a corrective disclosure. First, Plaintiffs do not clearly or specifically allege how this announcement corrected any specific previous misrepresentation or false statement made by the Opendoor Defendants. (See generally id. ¶¶ 220-24.) At most, it may have corrected the notion that Opendoor‘s algorithm was infallible; but even that conclusion assumes that the algorithm was responsible for the decrease in contribution margin. Regardless, as the Court has previously noted, the Opendoor Defendants’ statements would not have misled a reasonable investor into believing that its algorithm
ii. Bloomberg Article
On September 19, 2022, Bloomberg published an article revealing “that Opendoor lost money on 42 percent of its transactions in August 2022 (as measured by the prices at which it bought and sold properties).” (Doc. 39 ¶ 226.) Over the next two trading sessions, Opendoor‘s stock price fell by $0.50, closing at $3.56 on September 20, 2022. (Id. ¶ 229.) Again, Plaintiffs allege that the Opendoor Defendants’ misrepresentations regarding the efficacy of its algorithm were partially revealed. (Id. ¶ 228.) But the Bloomberg article revealed little other than that Opendoor‘s algorithm may be struggling with the sudden housing slowdown. (See id. ¶ 226.) This possibility had been forecasted by Opendoor‘s February announcement. Further, as Plaintiffs concede, all the data relied upon by Bloomberg was publicly available to investors. (Doc. 55 at 36.) Thus, the Court finds that the Bloomberg article does not constitute a corrective disclosure.
iii. Negative Contribution Margin
Finally, on November 3, 2022, Opendoor announced its financial results for the third quarter of 2022. (Doc. 39 ¶ 232.) Opendoor revealed a contribution margin of negative 0.7%. (Id.) On November 4, the price of Opendoor stock fell $0.32 and closed at $2.02. (Id. ¶ 235.) The next trading day, it fell by an additional $0.29 and closed at $1.74. (Id.) Plaintiffs allege that the November announcement revealed that “Opendoor‘s pricing algorithm could not adjust to changing housing markets or economic conditions and could not deliver 4% to 6% contribution margins in any market.” (Id. ¶ 234.)
Though it presents a closer question, the Court likewise finds that the November announcement was not a corrective disclosure. First, Opendoor had revealed months prior that its contribution margin was likely to fall significantly short of 4% for the third quarter.
Thus, for the additional and independent reason that Plaintiffs fail to plead loss causation, the Court finds that Plaintiffs’
B. Section 11
In case any part of the registration statement, when such part became effective, contained an untrue statement of a material fact or omitted to state a material fact required to be stated therein or necessary to make the statements therein not misleading, any person acquiring such security... may sue.
To prevail on a
1. Applicability of Rule 9(b)‘s Heightened Pleading Standard
Importantly, “the heightened pleading requirements of the
The parties disagree over whether
Plaintiffs counter that their
Plaintiffs’ disclaimer does not, on its own, evade
Plaintiffs’ cited authority does not compel an alternative conclusion. Knollenberg, 152 F. App‘x 674, In re Countrywide Financial Corp. Securities Litigation, 588 F. Supp. 2d 1132 (C.D. Cal. 2008) (hereinafter ”In re Countrywide“), and Tsirekidze v. Syntax-Brillian Corp., No. CV-07-02204-PHX-FJM, 2009 WL 275405 (D. Ariz. Feb. 4, 2009), each predate the Ninth Circuit‘s guidance in Rubke and In re Rigel. While the court in Boston Retirement System v. Uber Technologies, Inc., No. 19-cv-06361-RS, 2020 WL 4569846 (N.D. Cal. Aug. 7, 2020), held the plaintiffs’ disclaimer sufficient to evade
2. Actionable False or Misleading Statements
Defendants argue that Plaintiffs’
i. Statements Regarding the Pricing Algorithm‘s Efficacy and Role
Plaintiffs object to two statements made in Opendoor‘s offering documents: first, that Opendoor‘s “algorithms use machine learning to drive pricing decisions;” second, that the algorithm “can dynamically adjust to leading market indicators and react to real-time macro- and micro-economic conditions.” (Doc. 39 ¶¶ 402, 410; Doc. 56 at 15.) They contend that the statements “were false and misleading because they led investors to believe that advancements in [Opendoor‘s] algorithm drove [Opendoor‘s] pricing decisions and allowed Opendoor to ‘dynamically adjust’ to changing housing markets, when in reality, [Opendoor‘s] algorithm was already inaccurate and [Opendoor‘s] pricing decisions were driven by an undisclosed human-driven process.” (Doc. 56 at 17.)
The Court has already addressed whether these statements were false or misleading in the
ii. Statements Regarding the Pricing Algorithm‘s Importance
Plaintiffs next challenge the offering documents’ statement that Opendoor‘s
The Court also considered whether this statement was false or misleading in the
iii. Statements Regarding Opendoor‘s Business Model
Plaintiffs finally challenge the offering documents’ statement that Opendoor‘s “business model is designed to generate margins from [its] service charge to sellers and ancillary products and services associated with a transaction, and not from the spread between acquisition price and resale price.” (Doc. 39 ¶¶ 407, 414; Doc. 56 at 20-21.) Plaintiffs allege that the statement was false and misleading because, considered alongside “the other false and misleading statements in the offering documents, it led investors to believe that Opendoor‘s business model was disrupting the real estate market, when in reality, Opendoor‘s purportedly tech-based business was nothing more than a traditional real estate company that bought low and sold high.” (Doc. 56 at 21 (cleaned up).)
For the reasons articulated in the Court‘s discussion of the same challenged statement in the
3. Negative Causation
Plaintiffs’ claims under
A defendant raising the negative causation defense bears the burden of proof, and it is a “heavy burden.” Id. They “must show that the depreciation in value of a plaintiff‘s stock resulted from factors other than the alleged material misstatement.” Id. (internal quotations omitted). Additionally, “[b]ecause an analysis of causation is often fact-intensive, negative causation is generally established by a defendant on a motion for summary judgment or at trial.” In re Countrywide, 588 F. Supp. 2d at 1171.
As analyzed above, the Court finds that none of the purported corrective disclosures cited by Plaintiffs in their briefs withstand scrutiny. Courts within the Ninth Circuit and elsewhere have found negative causation established where the defendants have demonstrated that the plaintiffs failed to allege a corrective disclosure or any other indication of loss causation. See e.g., Brown v. Ambow Educ. Holding Ltd., No. CV 12-5062 PSG (AJWx), 2014 WL 523166, at *16 (C.D. Cal. Feb. 6, 2014) (collecting cases). And while, in some instances, it may be appropriate to reserve the issue for a later time, the Court concludes that postponing the issue would serve no legitimate purpose. See id. at *14 (“While the loss causation analysis is fact intensive and often more appropriate at the summary judgment phases, courts in the Ninth Circuit routinely apply
For these reasons, the Court finds that Plaintiffs’ claims against Defendants brought pursuant to
C. Sections 15 and 20(a)
A claim under
Similarly,
Because Plaintiffs fail to plead a primary violation of
IV. LEAVE TO AMEND
Plaintiffs “request leave to replead.” (Doc. 56 at 23 n.9; see also Doc. 55 at 38 n.17.) District courts “shall grant leave to amend freely ‘when justice so requires.‘” Lopez v. Smith, 203 F.3d 1122, 1130 (9th Cir. 2000) (en banc) (quoting
It is not clear to the Court that any further amendment to Plaintiffs’ Consolidated Amended Complaint would be futile. See Eminence Capital, 316 F.3d at 1052 (reasoning that dismissal with prejudice and “without leave to amend is not appropriate unless it is clear on de novo review that the complaint could not be saved by amendment“). Thus, Plaintiffs will be granted leave to file a Second Consolidated Amended Complaint.
Accordingly,
IT IS ORDERED that the Opendoor Defendants’ Motion to Dismiss and Memorandum of Points and Authorities in Support Thereof (Doc. 48) is granted.
IT IS FURTHER ORDERED that the Underwriter Defendants’ Motion to Dismiss the Consolidated Amended Complaint and Memorandum of Points and Authorities in Support Thereof (Doc. 51) is granted.
IT IS FINALLY ORDERED that Plaintiffs’ Consolidated Amended Complaint (Doc. 39) is dismissed with leave to amend. Plaintiffs may file a Second Consolidated Amended Complaint no later than 30 days from the date of this Order.
Dated this 27th day of February, 2024.
Michael T. Liburdi
United States District Judge