ORDER GRANTING IN PART AND DENYING IN PART DEFENDANT’S MOTION TO DISMISS [93]

I. INTRODUCTION

This is a putative class action for securities fraud under sections 10(b) and 20(a) of the Securities Exchange Act of 1934. Plaintiffs Noe Barocio, Salvador Barocio, and Cindy Conybear allege that Defendant Yingli Green Energy Holding Company *956Limited (“Yingli”), a company that sells solar energy products, defrauded its investors by making false and misleading public statements about (1) the company’s involvement in a government subsidy program, and (2) the collectability of debts owed by is customers. Yingli now moves to dismiss Plaintiffs’ Consolidated Amended Complaint, which the Court GRANTS IN PART and DENIES IN PART. (ECF No. 93.)1

II. FACTUAL BACKGROUND

Yingli is a publicly traded corporation that manufactures and sells solar energy products. (Consol. Am. Compl. (“CAC”) ¶ 2, ECF No. 92.) Prior to 2010, Yingli sold its products mainly to European companies. (Id. ¶¶ 3-4, 30-31.) Beginning in 2010, however, Yingli’s sales gradually shifted from Europe to China. This shift was due in part to a solar energy subsidy program offered by the Chinese government called “Golden Sun.” (Id. ¶¶ 5-9, 30-41.)

A. Golden Sun Program

Under the Golden Sun Program, the Chinese government subsidized up to 70% of the cost of approved solar power projects in China. (Id. ¶ 42.) Yingli benefitted from the program in two ways. First, the developers for one-quarter of all Golden Sun projects purchased photovoltaic panels from Yingli. (Id. ¶¶9, 48, 49.) Second, Yingli received subsidies for its own solar power projects. (Id. ¶ 49.)

1. The Allegedly Misleading Statements

Between December 2, 2010, and March 4, 2013, Yingli touted its involvement in— and attributed its success in the Chinese market to — the Golden Sun Program. (Id. ¶¶ 51-67.) For example, Yingli expressed “a firm commitment to the Golden Sun Program” and claimed to “have established a solid market position” through its involvement in the program. (Id. ¶¶ 52, 55.) Yingli attributed its “strong performance” in 2010 to “the steady growth in the rooftop segment under the Golden Sun Program.” (Id. ¶ 56.) During its Q3 2012 Earnings Call, Yingli announced that it expected “[t]he Golden Sun volume for next year [to] be much larger than this year,” and that “there will be no any [sic] cut.” (Id. 11.62.) And during its Q4 2012 Earnings Call, Yingli stated that “in the future, our profitability points are really coming from [the] Golden Sun Program.” (Id. ¶ 67.)

2. Why These Statements Were Misleading

According to Plaintiffs, these statements were misleading because Yingli failed to disclose two significant risks to its involvement in and reliance on the Golden Sun Program: (1) the inevitable termination of the program due to widespread fraud in obtaining subsidies; and (2) the government’s right to claw back subsidy awards from developers that did not meet project deadlines.

Fraud. Plaintiffs allege that at least 29% of Golden Sun subsidies were procured through “outright fraud,” making it all but inevitable that the government would terminate the program once the fraud came to light. (Id. ¶¶ 68, 77.) Such fraud included overstating project costs in subsidy applications, agreeing to use expensive high-quality materials but instead using inexpensive low-quality materials, and otherwise “submitting] false applications using fraudulent paperwork.” (Id. ¶77.) Yingli allegedly knew about the fraud because it procured subsidies for both itself and its *957customers through similar types of fraud. (Id. ¶ 109.) First, Yingli purposely overstated costs in both its own subsidy applications and the applications of its customers. For example, while Yingli sold solar panels at a “typical” rate of RMB 6 per watt, Yingli reported on its applications a “typical” rate of RMB 10 per watt. (Id. ¶ 110.) Second, Yingli deliberately delayed construction of its approved Golden Sun projects. (Id. ¶ 111.) The Chinese government would pay out subsidies immediately upon project approval, yet Yingli and its customers would delay construction until the market price for various project materials had decreased. (Id.) For example, the government approved one particular Yingli project in 2012 based on a cost estimate of RMB 13 per watt, but by the time construction on the project began, the cost had fallen to RMB 7-8 per watt. (Id.) This resulted in Yingli and its customers receiving an “interest free loan[ ].”2 (Id.) Finally, while Yingli was required to use certain high-quality materials, it instead used cheaper low-quality materials during actual construction. (Id. ¶ 112.) For example, Yingli represented that it would use 240-watt solar panels, but it instead used 235-watt panels for its projects. (Id.)

According to a former Yingli employee (FE2), the Chinese government discovered several instances of fraud in the program in 2009 and 2010. (Id. ¶ 113.) This prompted the government to require inspections of all projects approved after September 2010. (Id.) Yingli managed to avoid detection by showing government inspection teams only its compliant projects and convincing them that they need not inspect its non-compliant projects. (Id.)

Clawbacks. Plaintiffs also allege that there was a high risk that the government would claw back subsidies paid to both Yingli and its customers. The government notice announcing the Golden Sun Program stated that projects approved between 2009 and 2011 must be completed by February 15, 2012, and that the failure to meet this deadline would require the award recipient to repay the subsidy for that project. (Id. ¶¶ 9, 46.)3 Plaintiffs also allege that the widespread fraud in the Golden Sun Program exposed the subsidies to clawbacks, although it is unclear what types of misconduct would result in clawbacks. (See id. ¶¶ 82, 114.)

3. Clawbacks and Cancellation of the Golden Sun Program

Between March 18 and March 22, 2013, a series of news articles and several industry experts predicted that the Chinese government would discontinue the Golden Sun Program. (Id. ¶¶ 69-79.) At least one article noted that the program provided developers with “an overgenerous capital expenditure-based subsidy before installation,” thereby reducing their incentive to build high-quality solar energy systems. (Id. ¶ 69.) However, the article did not explicitly cite this as the reason why the government might discontinue the program. These articles and predictions allegedly caused Yingli’s stock price to fall 22.2% on March 25, 2013. (Id. ¶ 73.)

In April 2013, the Chinese Ministry of Finance issued clawback notices to developers that received subsidy awards between 2009 and 2011 but that had failed to complete their projects on time. (Id. ¶ 74.)4 On May 20, 2013, Sina.com, an ag-*958gregator of Chinese-language news, reported that the Ministry of Finance issued clawback notices to 109 Golden Sun projects, demanding a total repayment of between RMB 7 and 10 billion, {Id. ¶ 75.) Citing a Yingli “sales head,” the article further indicated that 51 of the 55 Golden Sun developers' that had purchased solar energy products from Yingli received claw-back notices, thus endangering “nearly 100 million” RMB in its accounts receivable. {Id. ¶ 87.) The article noted that the Golden Sun Program had been plagued by “continuous rumors” about project irregularities, such as “receiving subsidies by swindling, procrastinating on work schedules and passing substandard products as good products.”5 The article further noted that “problems have repeatedly occurred in terms of project examination and approval, subsidy disbursements and . subsequent supervision.”6

On June 8, 2013, the China Economic Weekly reported that the Ministry of Finance demanded the repayment of 80% of the subsidies that it awarded between 2009 and 2011. (Id. ¶ 76.) On June 20, 2013, the Chinese National Audit Office issued an Audit Report estimating that 29% of Golden . Sun subsidies awarded between 2009 and 2011 were “procured through intentional fraud,” (Id, ¶¶77, 78.) Five days later, an industry magazine cited the Audit Report as an indication that the Golden Sun Program was “nearing its end,” and that a new subsidy program “based on real power production” would take its place. (Id. ¶ 79.) Sure enough, the Chinese government cancelled the Golden Sun Program in December 2013. (Id. ¶¶ 81-82.)

B. Yingli’s Doubtful Accounts

Plaintiffs also allege that Yingli committed accounting fraud by delaying the recognition of doubtfúl accounts (i.e., accounts on which collectability is no longer reasonably assured) in the wake of Golden Sun’s collapse.

1. The Allegedly Misleading Statements

Yingli prepared its 2013 20-F Report in accordance with U.S. Generally Accepted Accounting Principles (GAAP). (Id. ¶85.) In this report, Yingli stated that it “establish[es] an allowance for doubtful accounts for the estimated loss on receivables when collection may no longer be reasonably assured.” (Id. ¶ 83.) Yingli then recognizes doubtful accounts as bad debt once “all means of collection have been exhausted and the potential for recovery is considered remote.” (Id. ¶ 84.)

2. Why These Statements Were Misleading

According to a former Yingli, employee (FE1), Yingli would delay making an allowance for. doubtful accounts until long after collection was no longer reasonably assured. That is, Yingli would not make such an allowance until (1) ‘Yingli ha[d] lost virtually all hope of collecting” the debt, and (2) Yingli had “obtained] permission from the tax office to write off income from the debts from Yingli’s taxes.” {Id. ¶ 88.) Thus, Yingli delayed recognizing , as doubtful the outstanding .accounts of its Golden Sun customers until 2014, even though the clawbacks rendered those accounts obviously uncollectible in 2013. {Id. ¶¶ 94-, 101.) Yingli’s 2013 20-F *959Report therefore misrepresented the company’s true financial condition.

Plaintiffs point to the account of Shanghai Chaori Solar Energy Science & Technology Co. Ltd. (“Chaori”) as an example of Yingli delaying the recognition of an obviously doubtful account. Chaori had received subsidies from the Golden Sun Program, and its “existence was imperiled by the Golden Sun clawbacks.” (Id. ¶ 89.) Chaori owed a Yingli subsidiary RMB 75.3 million as of May 2013.7 (Id. ¶ 91.) In March 2013, Chaori sent a letter to Yingli stating that it did not have the cash to pay the debt on time, and requested a payment extension until the end of 2013. (Id.) In April 2013, Yingli sued Chaori for the outstanding amount. (Id.) By July 2013, Chao-ri’s other creditors had sued Chaori for a total of RMB 1.906 billion. (Id. ¶ 90.) In September 2013, a Chinese court awarded Yingli the full RMB 75.3 million in outstanding debt. (Id. ¶ 91.) In March 2014, Chaori defaulted on its government-issued notes. (Id. ¶ 93.) In July 2014, Chaori’s creditors successfully petitioned a court to place Chaori into bankruptcy. (Id.) Chaori subsequently advised Yingli to pursue creditor’s rights with the bankruptcy court. (Id.)

Although Yingli recognized this debt as a doubtful account in its 2014 report, Plaintiffs allege that Yingli should have done so in its 2013 report. (Id. ¶¶ 94, 95.) Moreover, because Yingli made only a RMB 20 million allowance for doubtful accounts in 2013, and because Yingli’s outstanding accounts relating to the Golden Sun Program allegedly amounted to hundreds of millions of RMB, Plaintiffs infer that Yingli did not make any doubtful account allowance that year for outstanding debt owed by customers that were subject to Golden Sun clawbacks. (Id. ¶¶ 95, 96.) Plaintiffs infer that Yingli delayed making such an allowance until 2014, when it recorded RMB 228.8 million in doubtful accounts.' (Id. ¶ 96.)

On March 25, 2014, in response to this large disclosure of doubtful accounts, Yingli’s stock price fell 15%. (Id. ¶ 99.) On May 15, 2015, Yingli reported that it was' writing off USD $33.2 million (approximately RMB 230 million) in doubtful accounts. (Id. ¶¶ 100-01.) Plaintiffs contend that these accounts all became uncollectible because of clawbacks from the Golden Sun Program. (Id. ¶ 101.) The next trading day, Yingli’s' stock price fell 12.4%. (Id. ¶ 102.) The following day, Yingli’s stock price fell an additional 37%. (Id.)

C. Procedural History

On May 28, 2015, Kevin Knox filed this action. (ECF No. 1.) Three weeks later, Bhimsain Mangla filed a near-identical action. (See Compl., Mangla v. Yingli Green Energy Holding Co. Ltd., et al., No. 2:15—cv-04600-ODW (MRWx) (C.D.Cal. June 17, 2015).) The Court consolidated the two actions, appointed Noe Barocio and Salvador Barocio as lead plaintiffs, and The Rosen Law Firm as lead counsel. See Knox v. Yingli Green Energy Holding Co. Ltd., 136 F.Supp.3d 1159 (C.D. Cal. 2015). Plaintiffs subseqüeritly filed a Consolidated Complaint, which Yingli moved to dismiss. (ECF Nos. 63, 65, 74.) The Court granted in part and deníéd in part Yingli’s Motion. See Knox v. Yingli Green Energy Holding Co. Ltd., No. 215CV040030DWMRWX, 2016 WL 6609210, at *1 (C.D. Cal. May 10, 2016). Plaintiffs thereafter filed a Consoli*960dated Amended Complaint. (ECF No. 92.) Yingli has again moved to dismiss the complaint. (ECF No. 93.) That Motion is now before the Court for consideration.

III. LEGAL STANDARD

The court may dismiss a complaint for failure to plead sufficient facts to support a claim for relief. Fed. R. Civ. P. 12(b)(6); Balistreri v. Pacifica Police Dep’t, 901 F.2d 696, 699 (9th Cir. 1990). In a typical section 10(b) action, a plaintiff must plead and prove, among other things, (1) a material misrepresentation or omission by the defendant and (2) scienter. Stoneridge Inv. Partners, LLC v. Sci.-Atlanta, 552 U.S. 148, 157, 128 S.Ct. 761, 169 L.Ed.2d 627 (2008); 17 C.F.R. § 240.10b-5. The plaintiff must plead these elements in accordance with Federal Rule of Civil Procedure 9(b) and the Private Securities Litigation Reform Act of 1995 (“PSLRA”), In re VeriFone Holdings, Inc. Sec. Litig., 704 F.3d 694, 701 (9th Cir. 2012), although the Rule 9(b). analysis is “effectively subsumed” under the stricter PSLRA analysis, Miss. Pub. Emps. Ret. Sys. v. Boston Sci. Corp., 523 F.3d 75, 85 n.5 (1st Cir. 2008).

A. Material Misrepresentation

To establish the first element, “a plaintiff must show that the defendant made a statement that was misleading as to a material fact.” Matrixx Initiatives, Inc. v. Siracusano, 563 U.S. 27, 38, 131 S.Ct. 1309, 179 L.Ed.2d 398 (2011) (emphasis in original) (footnote and some internal quotation marks omitted). A statement containing an express falsehood is sufficient, but not necessary, to satisfy this element, for a statement may still be false or misleading if it omits a critical fact. Brody v. Transitional Hosps. Corp., 280 F.3d 997, 1006 (9th Cir. 2002). But to show fraud by omission, the company’s affirmative statement must be more than simply incomplete; it must “create an impression of a state of affairs that differs in a material way from the one that actually exists.” Id.; see also In re Cutera Sec. Litig., 610 F.3d 1103, 1109 (9th Cir. 2010). Moreover, there must be “a substantial likelihood that the disclosure of the omitted fact would have been viewed by the reasonable investor as having significantly altered the ‘total mix’ of information made available.” Matrixx Initiatives, 563 U.S. at 38, 131 S.Ct. 1309 (quoting Basic Inc. v. Levinson, 485 U.S. 224, 238, 108 S.Ct. 978, 99 L.Ed.2d 194 (1988)). “Pursuant to the PSLRA, a complaint must ‘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.’ ” Ronconi v. Larkin, 253 F.3d 423, 429 (9th Cir. 2001) (quoting 15 U.S.C. § 78u-4(b)(1)).

B. Scienter

“The complaint must also ‘state with particularity facts giving rise to a strong inference that the defendant acted with the required state of mind’ — that is, that he acted with intentionality or deliberate recklessness or, where the challenged act is a forward looking statement, with ‘actual knowledge ... that the statement was false or misleading.’ ” Ronconi, 253 F.3d at 429 (quoting 15 U.S.C. §§ 78u-4(b)(2)(A), 78u-5(c)(1)(B)(i)) (footnotes and some citations omitted). To determine whether the plaintiff has shown a “strong inference” of scienter, the court “must engage in a comparative evaluation; it must 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, 127 S.Ct. 2499, 168 L.Ed.2d 179 (2007). “An inference of fraudulent intent may be plausible, yet less *961cogent than other, nonculpable explanations for the defendant’s conduct. To qualify as 'strong’ ... an inference of scienter must be more than merely plausible or reasonable — it must be cogent and at least as compelling as any opposing inference of nonfraudulent intent.” Id. This analysis requires the court to “assess all the allegations holistically” rather than “scruti-niz[ing] each allegation in isolation.” Id. at 326, 127 S.Ct. 2499.

C. Purpose

The PSLRA’s strict pleading requirements were “[d]esigned to curb perceived abuses of the § 10(b) private action — ‘nuisance filings, targeting of deep-pocket defendants, vexatious discovery requests and manipulation by class action lawyers,’ ” Id. at 320, 127 S.Ct. 2499 (citation omitted). “Congress clearly intended that complaints in these securities actions should stand or fall based on the actual knowledge of the plaintiffs rather than information produced by the defendants after the action has been filed.” Medhekar v. U.S. Dist. Ct. for the N. Dist. of Cal., 99 F.3d 325, 328 (9th Cir. 1996). Thus, “[w]here pleadings are not sufficiently particularized or where, taken as a whole, they do not raise a ‘strong inference’ that misleading statements were knowingly or [with] deliberate recklessness made to investors, a private securities fraud complaint is properly dismissed under Rule 12(b)(6).” 15 U.S.C. § 78u-4(b)(3)(A).

D. Leave to Amend

Generally, the court should liberally grant the plaintiff leave to amend a dismissed complaint. Lopez v. Smith, 203 F.3d 1122, 1130 (9th Cir. 2000) (en banc). However, a court may deny leave to amend when it “determines that the allegation of other facts consistent with the challenged pleading could not possibly cure the deficiency,” Schreiber Distrib. Co. v. Serv-Well Furniture Co., 806 F.2d 1393, 1401 (9th Cir. 1986), or where the plaintiff has previously failed to cure pleading deficiencies identified by the court, Moore v. Kayport Package Exp., Inc., 885 F.2d 531, 538 (9th Cir. 1989).

IV. DISCUSSION

Plaintiffs assert three distinct theories of liability against Yingli. The first two relate to undisclosed risks concerning the Golden Sun Program. Plaintiffs allege that Yingli’s optimistic statements about the Golden Sun Program were misleading because Yingli failed to disclose the risk that the Chinese government (1) could claw-back subsidies for projects that were not finished on time, and (2) would likely discontinue the program due to widespread fraud in procuring subsidies. (CAC ¶¶ 51-82.) Under the third theory, Plaintiffs allege that Yingli’s statements regarding when it would recognize doubtful accounts wére false because Yingli delayed such recognition until long after collectability was no longer reasonably .assured. (CAC ¶¶ 83-102.) Yingli moves to dismiss each theory on several grounds. The Court addresses each in turn.

A. Undisclosed Risks in the Golden Sun Program

1. Non-Actionable Puffing

Yingli first argues that “most, if not all, of the Golden Sun statements are nonac-tionable puffery.” (Mot. at 9, ECF No. 93-1.) The Court agrees that some, but not all, of Yingli’s statements constitute puf-fery.

Puffing statements are optimistic statements by corporate executives that “are generally not capable of objective verification, and lack a standard against which a reasonable investor could expect them to be pegged.” Mulligan v. Impax Labs., Inc., 36 F.Supp.3d 942, 966 (N.D. Cal. *9622014) (citations and internal quotation marks omitted). “ ‘[Professional investors, and most amateur investors as well, know how to devalue the optimism of corporate executives,’ ” Cutera Sec. Litig., 610 F.3d at 1111 (quoting In re VeriFone Sec. Litig., 784 F.Supp. 1471, 1481 (N.D. Cal. 1992), aff'd, 11 F.3d 865 (9th Cir. 1993)), and thus such statements “are considered immaterial and [are] discounted by the market.” In re OmniVision Techs., Inc. Sec. Litig., 937 F.Supp.2d 1090, 1102 (N.D. Cal. 2013) (citations and internal quotation marks ómitted). As a result, puffery cannot give rise to a securities fraud claim. See, e.g., Oregon Pub. Emps. Ret. Fund v. Apollo Grp. Inc., 774 F.3d 598, 606 (9th Cir. 2014); Cutera Sec. Litig., 610 F.3d at 1111; Glen Holly Entm’t, Inc. v. Tektronix, Inc., 352 F.3d 367, 379 (9th Cir. 2003).8 That said, non-actionable puffing provides a basis for dismissal only where “the statement is ‘so obviously unimportant to a reasonable investor that reasonable minds could not differ on the question óf their unimportance.’ ” In re Energy Recovery Inc. Sec. Litig., No. 15-CV-00265-EMC, 2016 WL 324150, at *20 (N.D. Cal. Jan. 27, 2016) (quoting In re Ford Motor Co. Sec. Litig., 381 F.3d 563, 570 (6th Cir. 2004)).

The Court concludes that Yingli’s statements in paragraphs 52, 60, 61, and 65 of Plaintiffs’ CAC constitute non-actionable puffing.9. These are simply vaguely optimistic statements about the Golden Sun Program generally, and they do not sufficiently connect Yingli’s current or future success to the program. The remaining statements, however, expressed optimism about Yingli’s market position and potential for growth based specifically on the success of the Golden Sun Program. If, as Plaintiffs allege, Yingli knew that it had no hope of profiting from the program because clawbacks and discontinuation of the program were inevitable,- then a reasonable investor might conclude that the cheery statements “affirmatively create[d] an impression of a state of- affairs' that differs in a material way from the one that *963actually exists.” Brody, 280 F.3d at 1006. This is enough to make out a securities fraud claim at the pleading stage.10

2. Clawbacks

i. Falsity/Materiality

Yingli argues that Plaintiffs do not allege particular facts showing that the risk of clawbacks was material at the time Yingli made its optimistic statements about Golden Sun, (Mot. at 9-10.) Plaintiffs respond that Yingli’s knowledge of the fraud in the Golden Sun Program shows that Yingli should have known that potential clawbacks would affect its bottom line. (Opp’n at 16, ECF No. 99.) Plaintiffs also point to their particular allegations about how governmént clawbacks of Chaori’s subsidies endangered Yingli’s outstanding accounts. (Id. at 16.) The Court agrees with Yingli that Plaintiffs have not adequately demonstrated materiality.

Section 10(b) does not require that companies predict the future, and thus “[a] plaintiff may not plead ‘fraud by hindsight’; i.e., a complaint ‘may not simply contrast a defendant’s past optimism with less favorable actual results’ in support of a claim of securities fraud.” ACA Fin. Guar. Corp. v. Advest, Inc., 512 F.3d 46, 62 (1st Cir. 2008) (citation omitted); see also, e.g., In re GlenFed, Inc. Sec. Litig., 42 F.3d 1541, 1548 (9th Cir. 1994). However, section 10(b) does require that a company disclose the risk that a future event might occur if that risk is material. In re Nuvelo, Inc. Sec. Litig., 668 F.Supp.2d 1217, 1230 (N.D. Cal. 2009) (failure to disclose then-existing risk does not constitute fraud-by-hindsight); Knox, 2016 WL 6609210, at *8 (same).

One of the undisclosed risks here was the risk that the government would claw-back subsidies if developers did not complete their projects on time, as was their right under the subsidy contracts. The materiality of this risk to Yingli depended on two main factors: (1) the likelihood that Yingli’s customers would not timely complete their projects; and (2) Yingli’s potential exposure in the event of clawbacks. Logic dictates that while more of one requires less of the other, at least some showing of each is required to demonstrate materiality. Further, the Court must consider them as of the time Yingli made its allegedly misleading statements.

Plaintiffs entirely fail to establish the first factor, because neither the fraud in the Golden Sun Program nor the Chaori receivable show any likelihood that Yingli’s customers would not complete their projects on time. The fraud in the program is irrelevant because Plaintiffs base their clawback theory on the failure of Yingli customers to meet project deadlines, not on the fraudulent procurement of subsidies.11 Yingli’s purported involvement in *964such fraud says nothing about the likelihood that its customers would not complete their projects on time. The Chaori receivable is also insufficient, because Plaintiffs’ assertions that Chaori was connected to Golden Sun and that its existence was “imperiled” by the clawbacks are too conclusory and unparticularized. (CAC ¶ 89.) And even if the Court took these assertions at face value, they say nothing of the likelihood that Chaori would not meet its project deadlines as of the time Yingli issued its misleading statements.12 Thus, any reliance on the Chaori allegations would be fraud by hindsight.

In sum, Plaintiffs have not presented particular facts in existence at the time of Yingli’s optimistic statements showing any likelihood that its customers would not meet their project deadlines. Without this, there is nothing to show that potential clawbacks presented a material risk to Yingli’s involvement in Golden Sun at that time. Ergo, Yingli need not have disclosed that risk.

ii. Scienter

Yingli argues that Plaintiffs have failed to establish that any particular Ying-li executive or director knew about the risk of clawbacks. (Mot. at 12-13.) In response, Plaintiffs rely on the “absurd to suggest” doctrine: that the risk of clawbacks was so obvious that it would be absurd to suggest that the risk was not known to Yingli’s upper management or that the failure to disclose the risk was an innocent error in judgment. See Berson v. Applied Signal Tech., Inc., 527 F.3d 982, 989 (9th Cir. 2008). Plaintiffs’ argument fails. As noted above, Plaintiffs do not plead facts showing that there was even a material risk of clawbacks at the time Yingli made its optimistic statements about Golden Sun. Thus, the allegations obviously cannot show that the risk was so apparent that Yingli’s executives could not possibly have been unaware of it. See Knox, 2016 WL 6609210, at *13.

iii. Leave to Amend

While the court should liberally grant leave to amend, it need not do so if the plaintiff has failed to cure pleading deficiencies previously noted by the Court. See Moore, 885 F.2d at 538. Here, the Court held that Plaintiffs’ prior iteration of the complaint failed to show that the risk of clawbacks was sufficiently material to require disclosure, or that Yingli acted with scienter. Knox, 2016 WL 6609210, at *9-10. Plaintiffs’ amended complaint does virtually nothing to cure these deficiencies. The Court therefore dismisses this theory without leave to amend.

8. Subsidy Fraud/Cancellation of Golden Sun

i. Falsity/Materiality

Yingli argues that because the Chinese government had not given any indication it was going to discontinue the Golden Sun Program when Yingli made its optimistic statements, Plaintiffs’ theory is simply fraud by hindsight. (Mot. at 13-14.) Plaintiffs do not dispute that Yingli had no actual knowledge that the government would ultimately cancel Golden Sun; rather, Plaintiffs argue that there was always a material risk that the government would eventually take some sort of drastic measure once it discovered the scale of the fraud. (Opp’n at 14-15.)

The Court agrees with Plaintiffs. The fact that the Chinese government had not made public its intent to cancel the Golden *965Sun Program is not dispositive of the falsity of Yingli’s statements. Instead, whether the potential cancellation of Golden Sun was a material risk to Yingli depended on three factors: (1) the likelihood that the government would discover the fraud; (2) the likelihood that the government would cancel the program upon such discovery; and (3) the potential effect on Yingli’s bottom line if it did so. Again, these factors operate on a sliding scale, and must be considered as of the time Yingli made its allegedly misleading statements. See GlenFed, Inc. Sec. Litig., 42 F.3d at 1548 (no fraud by hindsight).

First, there appeared to be a reasonable likelihood that the government would discover the fraud. According to a former Yingli employee (FE2), the Chinese government discovered several instances of fraud in the program in 2009 and 2010, prompting them to inspect all projects approved after September 2010. (Id. ¶ 113.) Because the government discovered actual instances of fraud and put in place a mechanism to detect any further instances of fraud, there was a reasonable likelihood that the government would eventually discover the full extent of the alleged fraud in the program.

Yingli argues that the Court should not rely on allegations attributed to FE2. (See Mot. at 16-17.) Where the plaintiff relies on information provided by a confidential witness, that witness must be “described in the complaint with sufficient particularity to support the probability that a person in the position occupied by the source would possess the information alleged.” In re Daou Sys., Inc., 411 F.3d 1006, 1015 (9th Cir. 2005). And even where this threshold requirement is satisfied, the court is not required to take every statement of that witness at face value — the court should still evaluate “ ‘the level of detail provided by the confidential sources, the corroborative nature of the other facts alleged (including from other sources), the coherence and plausibility of the allegations, the number of sources, the reliability of the sources, and similar indicia.’” Id. (quoting In re Cabletron Sys., Inc., 311 F.3d 11, 29-30 (1st Cir. 2002)). Here, Plaintiffs describe FE2 as follows:

[FE2] was a head of project development at Yingli from December 2010 to May 2013. FE2’s principal responsibilities included Golden Sun project management, where he was mainly in charge of the project application process, including all interactions with government. FE2’s Golden Sun experience included Yingli Hunan Yueyang, a 20 MW project which received Golden Sun subsidies. FE2 reported to Yingli’s senior managers.

(CAC ¶ 28.)

Because FE2 was “in charge of the [Golden Sun] project application process, including all interactions with government,” the Court concludes that there is a sufficient probability that he would possess information relating to the Chinese government’s inspection requirements and the fact that the government was aware of instances of fraud between 2009 and 2010. (See id.) Moreover, the government’s June 2013 audit report regarding the extensive fraud tends to corroborate the government’s prior knowledge of the fraud (i.e., at the time Yingli made its optimistic Golden Sun statements), for the discovery of such wide scale fraud does not typically occur overnight. Cf. Berson, 527 F.3d at 988 n.5 (court may infer that one who discloses a fact publicly at a certain time knew about that fact at an earlier point).

Second, the Court agrees with Plaintiffs that there was a significant likelihood that the Chinese government would cancel the Golden Sun Program upon discovering such fraud. The scale of the fraud appeared to be large — Plaintiffs allege that *96629% of projects awarded between 2009 and 2011 were procured through some sort of fraud, involving over RMB 207 million. (CAC ¶ 77.) Moreover, the program inherently encouraged fraud by offering upfront subsidies based on estimates of project costs that were easy to inflate. Given the scale of the fraud and the fact that the program’s structure itself encouraged such fraud, there was a real chance that the government would cancel the program once, the scope of the fraud became known.

Finally, while Plaintiffs do not quantify the effect that such cancellation would have on Yingli’s bottom line, this is not fatal to establishing, materiality. Yingli issued multiple statements attributing its success in the Chinese market to the Golden Sun Program and predicting that its future success would come from the program. (See, e.g., CAC ¶ 67 (stating in March 2013, only months before Golden Sun’s collapse, that “our profitability points are really coming from Golden Sun Program”).) These statements are sufficient for the Court to infer that any discontinuation would have had a materially negative effect on Yingli’s future performance.

For these reasons, the Court concludes that Plaintiffs have sufficiently alleged that the failure to warn of the risk of termination of the Golden Sun Program constituted a material omission.

ii. Sciénter

Yingli makes four broad arguments as to why Plaintiffs’ allegations regarding scienter are insufficient. First, the examples that Plaintiffs provide regarding Yingli’s own fraudulent conduct — which supposedly show Yingli’s knowledge of industry-wide fraud — are insufficiently particularized. (Mot. at 14-16.) Second, FE2’s position at the company does not suggest that he knows the facts attributed to' him, and Plaintiffs do not corroborate those facts. (Id. at 16-17.) Third, Plaintiffs do not allege what Yingli’s executives knew (as opposed to what Yingli as" a company knew). (Id. at 17-18.) Finally, even if Ying-li’s executives knew about the company’s own fraudulent activities, there are no allegations showing that Yingli knew that the fraud was industry-wide. (Id. at 18.) In response, Plaintiffs rely on the “absurd to suggest” exception to the core operations theory — i.e., that Golden Sun was so important to Yingli, and Yingli was so involved in the fraud, that it would be absurd to suggest that Yingli did not know of it. (Opp’n at 16-20.) The Court agrees with Yingli that Plaintiffs have not adequately alleged scienter.

The core operations theory posits that “ ‘facts critical to a business’s core operations or an important transaction generally are so apparent that their knowledge may be attributed to the company and its key officers.’ ” S. Ferry LP, No. 2 v. Killinger, 542 F.3d 776, 783 (9th Cir. 2008); see also In re Read-Rite Corp., 335 F.3d 843, 848 (9th Cir. 2003). In Read-Rite, and again in Killinger, the Ninth Circuit held that a securities fraud plaintiff cannot “rely[ ] exclusively on the core operations inference to plead scienter under the PSLRA.” Killinger, 542 F.3d at 784. The only exception is the “rare” instance where “the nature of the relevant fact is of such prominehce that it would be ‘absurd’ to suggest that management was without knowledge of the matter.” Id. at 786; see also Berson, 527 F.3d at 989; No. 84 Emp’r-Teamster Joint Council Pension Tr. Fund v. Am. W. Holding Corp., 320 F.3d 920, 943 n.21 (9th Cir. 2003).

Plaintiffs do not meet this demanding standard. First, the allegations are insufficient to show that Yingli’s upper management knew of even its own alleged fraud, let alone industry-wide fraud. That is, while Plaintiffs allege that Yingli overstated costs on its Golden Sun applications *967and purposely delayed construction of projects, they do not mention on how many Yingli projects this occurred. If there are no facts showing, how prominent these frauds were, the Court obviously cannot infer that they were so prominent that management must have been aware of it. Similarly, while Plaintiffs assert that Ying-li’s substitution of 240-watt solar panels for cheaper 235-watt solar panels was “widespread,” happened “generally,” and “was the rule, not the exception” (CAC ¶ 112; Opp’n at 18), these vague quantifiers are insufficient. The Court cannot draw any meaningful inferences about what Yingli’s executives knew regarding this fraud based on an anonymous Yingli employee’s subjective belief that it was “widespread” or, worse, Plaintiffs’ counsel’s bare assertion that it occurred “generally” and “was the rule, not the exception.” 13 The Ninth Circuit cases relying on the “absurd to suggest” doctrine are usually based on concrete numbers, not majestic generalities. See Berson, 527 F.3d at 988 n.5 (pointing to two specific stop-work orders that halted between $18 and $23 riiil-lion of work from government agencies that provided 80% of the company’s revenue); America West, 320 F.3d at 928, 943 n.21 (pointing to specific stock repurchases worth $100 million and an FAA investigation in which the agency was contemplating $11 million in penalties).14 Second, there are insufficient facts from which to infer that Yingli’s upper management knew anything about íraud in the program generally. Even if (as Plaintiffs’ contend) Yingli’s customers constituted a large enough portion of the program that their frauds alone would be sufficient to shut it down, Plaintiffs give no facts regarding how frequently it assisted its customers in committing fraud. Thus, again, Plaintiffs do not show that it happened so often that Yingli’s upper management - must have known about it.

iii. Leave to Amend

In the Court’s order granting Yingli’s prior Motion to Dismiss, the Court dismissed Plaintiffs’ fraud theory with leave to amend on statute of limitations grounds, but it also noted that Plaintiffs’ allegations in any event “did not ... give rise to a strong inference of scienter.” Knox, 2016 WL 6609210, at *7. Here, although Plaintiffs’ allegations do not cure this deficiency, they come substantially closer to establishing scienter than they did before. This suggests that additional facts may still exist that Plaintiffs could plead that would establish a securities fraud claim, and thus that granting leave to amend would not be a futile endeavor. The Court will therefore grant leave to amend as to this theory.

4. Statute of Limitations

Yingli argues that Plaintiffs’ undisclosed-risk theories are barred by the statute of limitations., (Mot. at 18-19.) The Court declines to address this argument at this time. As both the Second and Third Circuits have held, “[o]nly after a plaintiff *968can adequately plead his claim can that claim be said to have accrued, and only after a claim has accrued can the statute of limitations on that claim begin to run.” City of Pontiac Gen. Emps.’ Ret. Sys. v. MBIA, Inc., 637 F.3d 169, 175 (2d Cir. 2011); see also Pension Trust Fund for Operating Eng’rs v. Mort. Asset Securitization Transactions, Inc., 730 F.3d 263, 275 (3d Cir. 2013) (following MBIA). Here, because Plaintiffs have failed to adequately plead their undisclosed-risk theories, Court cannot discern by what date these theories accrued. The Court therefore defers ruling on the statute of limitations issue until Plaintiffs adequately state a claim for relief.

B. Accounting Fraud: Recognition of Doubtful Accounts

1. Falsity

With respect to Plaintiffs’ accounting fraud theory, Yingli argues that Plaintiffs have not established falsity or materiality because: (1) the description in the Sina. com article regarding the collectability Yingli’s outstanding receivables is too con-clusory; (2) allegations attributed to FE1 concerning Yingli’s fraudulent accounting practices are insufficiently particularized and are not adequately corroborated; and (3) in any event the facts pleaded show that Yingli reasonably accounted for the Chaori receivable. (Mot. at 19-22.)

“To properly state a claim for accounting fraud, plaintiffs must plead facts sufficient to support a conclusion that defendant prepared the fraudulent financial statements and that the alleged financial fraud was material.” Daou Sys., 411 F.3d at 1016 (citations, brackets, and internal quotation marks omitted). Plaintiffs should generally include “(1) such basic details as the approximate amount by which revenues and earnings were overstated; (2) the products involved in the contingent transaction; (3) the dates of any of the transactions; or (4) the identities of any of the customers or company employees involved in the transactions.” Id. (brackets and internal quotation marks omitted). While the amount of detail required will vary from case to case, the bottom line is that plaintiffs “must allege enough information so that a court can discern whether the alleged GAAP violations were minor or technical in nature, or whether they constituted widespread and significant inflation of revenue.” Id. at 1017 (internal quotation marks omitted); see also id. at 1018 (“[T]he plaintiff must show with particularity how the adjustments affected the company's financial statements and whether they were material in light of the company’s overall financial position.”); Rieckborn v. Jefferies LLC, 81 F.Supp.3d 902, 928 (N.D. Cal. 2015) (holding that the plaintiff should identify, “specific accounts that were in jeopardy when the alleged misrepresentations were made, specific accounts in existence at the time the alleged misrepresentations were made that were ultimately rendered uncollectible, and when and to what extent [the company’s] reserves should have been changed”).

“Understatements of bad debt reserves can support a securities fraud claim because ‘[companies are obliged to make reasonable predictions about the col-lectability of their accounts receivable. Underestimates of bad debt reserves lead to overstatement of income, and ultimately inflation of stock price.’ ” Alaska Elec. Pension Fund v. Adecco S.A., 371 F.Supp.2d 1203, 1213 (S.D. Cal. 2005) (quoting Kane v. Madge Networks N.V., No. C-96-20652-RMW, 2000 WL 33208116, at *5 (N.D. Cal. May 26, 2000)). That said, doubtful account recognition is an imprecise science, because the underlying accounting concept — i.e., reasonable assurance of collectability — “is a matter of judgment and estimate.” In re Galileo *969Corp. S’holders Litig., 127 F.Supp.2d 251, 265 (D. Mass. 2001); cf. GlenFed, Inc. Sec. Litig., 42 F.3d at 1549 (“[T]he setting of loan loss reserves [is] based on flexible accounting concepts, which, when applied, do not always (or perhaps ever) yield a single correct figure.”). Thus, it is not enough for the plaintiff to show that the company was ultimately unable to collect on an account; this would simply be fraud by hindsight. Instead, the plaintiff must show that “no reasonable accountant would have made the same decision if confronted with the same facts” at the time the report was issued. Alaska Elec. Pension Fund, 371 F.Supp.2d at 1213 (citations and internal quotation marks omitted).

Plaintiffs’ accounting fraud theory goes like this: (1) the Chinese government clawed back hundreds of millions of RMB in Golden Sun subsidies from Yingli’s customers in 2013; (2) it was obvious before the end of the year that the clawbacks rendered Yingli’s customers’ debts uncollectible, and thus Yingli should have recognized these debts as doubtful accounts in 2013; and (3) Yingli delayed doing so until 2014, thereby committing accounting fraud. However, the facts alleged do not support many of the tenuous inferences Plaintiffs draw at the first two steps.

The First Step. Plaintiffs rely solely on the Sina.com article and Yingli’s 2014 financial statements to show that the Chinese government clawed back hundreds of millions of RMB in subsidies from Yingli’s customers in 2013. (Opp’n at 22; CAC ¶ 98.) Neither shows this. As to the Sina.com article, Plaintiffs must show that the article (1) is based on an independent investigative effort, (2) is sufficiently particular and detailed to indicate its reliability, and (3) is corroborated by Plaintiffs’ counsel’s own independent investigation. See In re JPMorgan Chase & Co. Sec. Litig., No. 06C4675, 2007 WL 4531794, at *5 (N.D. Ill. Dec. 18, 2007). Admittedly, the specific details in the Sina.com article give it a modicum of credibility: it specifies that 109 projects received clawback notices (amounting to between 7 billion and 10 billion RMB), and that 51 of Yingli’s 55 Golden Sun customers received clawback notices (endangering nearly RMB 100 million of Yingli’s receivables). However, Plaintiffs’ allegations do not independently corroborate these facts. Plaintiffs argue that their independent allegations regarding the fraud in the Golden Sun Program corroborate the article, but as already stated ad nauseam, Plaintiffs do not sufficiently allege that the fraud was connected with the clawbacks. Thus, these allegations do not corroborate the Sina.com article in any relevant manner.

As to Yingli’s financial statements, Plaintiffs contend that the RMB 228 million in doubtful accounts that Yingli recognized in 2014 all relate to Golden Sun subsidies and clawbacks, and thus Yingli should have recognized all of it as doubtful in 2013 after the clawbacks occurred. But it is simply Plaintiffs’ speculation that any of the 2014 write off was connected to clawbacks or even to Golden Sun at all; there are absolutely no particular allegations to this effect. The only sum of money that Plaintiffs even attempt to specifically link to Golden Sun is the RMB 75 million Chaori receivable, where Plaintiffs state simply that “Chaori was heavily involved in Golden Sun” and that “Chaori was one of the Yingli debtors whose existence was imperiled by the Golden Sun clawbacks.” (CAC ¶ 89.) This is not enough. Plaintiffs do not say how much money the Chinese government attempted to clawback from Chaori or how such clawbacks materially contributed to Chaori’s inability to repay its debts. Moreover, there is no reason for the Court to assume more broadly that every single Yingli debtor must have been involved in Golden Sun and that clawbacks *970must have crippled each of their ability to pay their debts. Thus, to the extent Plaintiffs use Chaori to show a connection between Golden Sun ,and the accounting fraud, it fails.

The Second Step. Plaintiffs also give scant facts regarding the likelihood of Yingli collecting on debts owed by its Golden Sun customers. To the extent Plaintiffs rely on the Sina.com article to show that Yingli should have immediately written off all such debts, it is again insufficient. The article relies on an anonymous Yingli sales head for the proposition that the claw-backs endangered up to RMB 100 million in Yingli’s accounts receivables. First, as the Court noted in its prior order, the mere fact of clawbacks is insufficient to conclude that “Yingli’s customers [were] incapable of paying their debts. Rather, its customers’ ability to satisfy their accounts with Yingli in the event of clawbacks depended] on, among other things, the customer’s financial health at that time, and the safeguards Yingli may have put in place to secure the debt.” Knox, 2016 WL 6609210, at *9. The Sina.com article does not discuss these issues. Second, it is unclear how a sales head would have either adequate training in accounting principles or sufficient knowledge regarding each customers’ financial health to make a reliable evaluation of their ability to repay their debts. Cf. Daou Sys., Inc., 411 F.3d at 1015 (confidential witnesses must be identified “with sufficient particularity to support the probability that a person in the position occupied by the source would possess the information alleged”); Police Ret. Sys. of St. Louis v. Intuitive Surgical, Inc., 759 F.3d 1051, 1063 (9th Cir. 2014) (court should not rely on the “impression[s]” of a witness that lacks “substance or context”).

Plaintiffs’ attempt to show that the RMB 75 million Chaori receivable was obviously uncollectible in 2013 does .not fare much better,15 for it is not clear that no reasonable accountant would have delayed recognizing it as a doubtful account until 2014. Chaori did not owe-Yingli this money until May 2013. While Chaori indicated in March 2013 that it could not meet this deadline, it suggested that it could repay this money by the end of the year. (CAC ¶ 91.) In September 2013, Yingli received a court judgment for the full amount of the debt. (Id.) Thus, collection on this debt was not necessarily doubtful as long as Yingli reasonably assumed that it could enforce this judgment. Plaintiffs suggest that Yingli could not have assumed this because creditors had already sued Chaori for a total of RMB 1.906 billion by July 2013. (Id. ¶90.) Not only do Plaintiffs fail to allege that Yingli knew this, but this number in itself is of limited value. First, the fact that creditors are seeking a certain amount of money does not necessarily mean that Chaori owes that amount of money. Second, unless Chaori’s outstanding debts exceeded its assets, Yingli could still reasonably assume that it could recover the full judgment against Chaori.16 As a result, Plaintiffs have not shown that any reasonable accountant would have written *971off the full RMB 75 million in 2013.17 See Alaska Elec. Pension Fund, 371 F.Supp.2d at 1213.

Finally, Plaintiffs make several other unsubstantiated allegations about Yingli’s accounting practices. First, Plaintiffs allege that Chaori owed Yingli 100 million RMB as of December 2012. This debt appears to be independent of the RMB 75 million receivable, given that the latter debt was not due until May 2013. Yet Plaintiffs never make any specific allegations as to when Yingli should have recognized this debt as doubtful, or when Yingli actually wrote it off (if ever). See Rieckborn, 81 F.Supp.3d at 928. Second, Plaintiffs allege that Yingli would delay recognition of bad debt until they secured a tax write-off for it. However, Plaintiffs do not connect this purported policy to specific accounts or dollar amounts that Yingli wrote off, and thus it is unclear how this practice ultimately affected Yingli’s accounting. Indeed, as Plaintiffs’ accounting fraud theory relates to doubtful account recognition and not bad debt recognition, it is unclear how a policy relating to bad debt recognition could have anything to do with Plaintiffs’ claims.

2. Materiality

Yingli also argues that the Chaori receivable is in any event immaterial because it made up only 0.27% of Ying-li’s overall assets. (Mot. at 23.) Plaintiffs argue that the proper- comparator is not total assets' but net losses, and that, the Chaori receivable made up 3.6% of net losses. (Opp’n at 21-22.) Generally, materiality is determined by comparing the. disputed accounting items “to like items on the corporate financial statement.” See Ganino v. Citizens Utils. Co., 228 F.3d 154, 165 (2d Cir. 2000). If the disputed items account for more than 5% of the “like items,” there is a presumption that the disputed items are material. Id.; see also IBEW Local Union No. 58 Pension Tr. Fund & Annuity Fund v. Royal Bank of Scotland Grp., PLC, 783 F.3d 383, 390 (2d Cir. 2015). But courts should also take into account qualitative factors in assessing materiality. Ganino, 228 F.3d at 165. Such factors may include: “whether the misstatement ‘arises from an item capable of precise measurement’; ‘masks a change in earnings or other trends’; ‘changes a loss into income or vice versa’; ‘concerns a segment or other portion of the ... business that has been identified as playing a significant role in the registrant’s operations or profitability’; ‘involves concealment of an unlawful transaction’; and whether ‘a known misstatement may result in a significant positive or negative market reaction.’ ” IBEW, 783 F.3d at 391 (quoting 64 Fed. Reg. 45,150, 45,151 (Aug. 19, 1999)).

*972The Court agrees with Plaintiffs that a reasonable investor may have found the RMB 75 million Chaori receivable material. “Under GAAP principles, doubtful accounts receivable — -where losses due to uncollectible receivables are probable and reasonably estimable — are to be charged to income.” In re: Ebix, Inc. Sec. Litig., 898 F.Supp.2d 1325, 1331 (N.D. Ga. 2012). A reasonable investor evaluates company losses not only in terms of its effect on total assets, but also for what it may signal about the future of the company — such as future profitability in key markets.18 Here, as Plaintiffs note, Yingli’s business was undergoing a major shift from Europe to China between 2010 and 2013. (CAC ¶¶ 4-7, 37.) As China was Yingli’s up-and-coming market, a reasonable investor may have viewed the recognition of a Chinese account as doubtful to be important despite its negligible impact on Yingli’s total assets. Thus, the fact that the Chaori receivable made up 3.6% of total losses — and a significantly larger percentage of losses within the Chinese market — may have caused a reasonable investor to find the Chaori receivable material.19

3. Scienter

Yingli argues that Plaintiffs have not established scienter because there are no allegations concerning what Yingli’s top executives knew. (Mot. at 23-25.) In response, Plaintiffs ask the Court to infer scienter based on: (1) Yingli’s policy of obtaining a tax write off before recognizing a doubtful account; (2) FEl’s “careful[] tracking]” of Yingli’s accounts receivable; and (3) the fact that Yingli’s Chief Executive Officer was named on the court opinion awarding Yingli the full amount of the RMB 75 million Chaori debt. (Opp’n at 23-25.)

Generally, “ ‘the mere publication of inaccurate accounting figures, or a failure to follow GAAP, without more, does not establish scienter.’ ” DSAM Glob. Value Fund v. Altris Software, Inc., 288 F.3d 385, 390 (9th Cir. 2002) (quoting In re Software Toolworks Inc., 50 F.3d 615, 627 (9th Cir. 1994)); see also In re Worlds of Wonder Sec. Litig., 35 F.3d 1407, 1426 (9th *973Cir. 1994); Provenz v. Miller, 102 F.3d 1478, 1490 (9th Cir. 1996); Malone v. Microdyne Corp., 26 F.3d 471, 479 (4th Cir. 1994); In re Northpoint Commc’ns Grp., Inc. Sec. Litig., 184 F.Supp.2d 991, 998 (N.D. Cal. 2001). This is particularly so in the realm of doubtful accounts or bad debt recognition due to the elasticity of the underlying accounting principles. See Coates v. Heartland Wireless Commc’ns, Inc., 55 F.Supp.2d 628, 640 (N.D. Tex. 1999) (“Questioning the timing of the writedown ... does not provide a strong inference of fraud.”); DiLeo v. Ernst & Young, 901 F.2d 624, 627 (7th Cir. 1990) (“For any bad loan the time comes when the debtor’s failure is so plain that the loan is written down or written off. No matter when a bank does this, someone may say that it should have acted sooner. If all that is involved is a dispute about the timing of the writeoff, based on estimates of the probability that a particular debtor will pay, we do not have fraud; we may not even have negligence.”); Louisiana Sch. Emps. Ret. Sys. v. Ernst & Young, LLP, 622 F.3d 471, 482 (6th Cir. 2010); cf. In re Ikon Office Sols., Inc., 277 F.3d 658, 675 (3d Cir. 2002) (“[A]uditing estimated reserves for doubtful accounts is a highly imperfect undertaking that requires an assessment of the risk that accounts may be defaulted on. As there is no evidence to suggest that Ernst’s method of predicting colleetibilty was unreasonable or grossly inconsistent with acceptable accounting practices, there is no basis to conclude that Ernst fraudulently certified that the reserve for doubtful accounts in IKON’s consolidated financial statements comported with GAAP.”).

None of the facts Plaintiffs point to give rise to a strong inference of scienter. First, as previously noted, Plaintiffs do not show that Yingli’s purported policy of obtaining a tax write-off before recognizing an account as doubtful affected a particular account or even a particular sum of money. Rather, this appears to be simply a stray allegation of wrongdoing, which does not help establish scienter. Knox, 2016 WL 6609210, at *16; Zucco Partners, LLC v. Digimarc Corp., 552 F.3d 981, 1008 (9th Cir. 2009) (holding that “a large quantity of otherwise questionable allegations” does not create a strong inference of scienter); Bricklayers of W. Penn. Pension Plan v. Hecla Min. Co., No. 2:12-CV-00042-BLW, 2013 WL 5423875, at *8 (D. Idaho Sept. 26, 2013) (“unrelated and irrelevant” allegations of misconduct do not help to establish scienter). Second, FE1 was an accountant at a Yingli subsidiary, not at Yingli. It thus unclear how FE1 would have personal knowledge regarding Yingli’s accounting practices. And while the Court can consider facts known to a confidential witness through hearsay, Lloyd v. CVB Fin. Corp., 811 F.3d 1200, 1208 (9th Cir. 2016), Plaintiffs do not say how FE1 learned of Ying-li’s accounting practices, or that he relayed this information to Yingli’s upper management. The Court also notes that while FE1 makes the general assertion that Yingli conducted monthly account reconciliations with its customers, Plaintiffs do not identify specific meetings, customers, or accounts where this happened — including, most importantly, the Chaori account. Third, the fact that Yingli’s CEO is listed on the court opinion regarding the Chaori receivable also does not shovv scienter. The facts asserted in the Chinese court’s opinion regarding the Chaori account does not show that Yingli should have recognized the account in 2013, and thus those facts do nothing to show that Yingli CEO’s acted with fraudulent intent.

4. Leave to Amend

The Court previously held that Yingli’s accounting fraud theory, which at that time was based on improper recognition of revenue rather than delayed recognition of doubtful accounts, did not allege with par*974ticularity either a misrepresentation or scienter. Knox, 2016 WL 6609210, at *11-13. Despite the shift in theory, Plaintiffs do not allege nearly sufficient facts to demonstrate widespread accounting irregularities relating to the Golden Sun Program. However, Plaintiffs come closer to asserting an accounting fraud theory based on the RMB 75 million Chaori receivable. The Court therefore grants leave to amend to assert an accounting fraud theory based on the Chaori receivable only. The accounting fraud theory is otherwise dismissed without leave to amend.

V. CONCLUSION

For the foregoing reasons, the Court GRANTS IN PART and DENIES IN PART Yingli’s Motion to Dismiss. (ECF No. 93.) To the extent that the .Court grants leave to amend, Plaintiffs may file a further amended complaint' addressing the deficiencies noted in this Order no later than April 24, 2017. Yingli shall have 35 days to respond to the complaint. If Ying-li’s response is a pre-answer motion, Plaintiffs shall file their opposition within 35 days after the motion is filed, and Yingli may file a reply within 14 days thereafter. Yingli should set the motion for hearing on August 14, 2017, at 1:30 p.m.

IT IS SO ORDERED.