OPINION OF THE COURT
ALITO, Circuit Judge:
Frederick F. Keller sued his former employer, ORIX Credit Alliance, Inc., in federal district court, asserting claims under the federal Age Discrimination in Employment Act (“ADEA”), 29 U.S.C. § 621 et seq., and the [1103]*1103New Jersey Law Against Discrimination (“NJLAD”), N.J.SA. § 10:5-1 et seq. Keller, who had served as executive vice president and a member of the board of directors, claimed that ORIX Credit Alliance had discriminated against him based on his age when it failed to promote him to the position of chief operating officer and later terminated his employment. The district court granted summary judgment for ORIX Credit Alliance. A panel of this court issued a decision reversing the district court, but ORIX Credit Alliance’s petition for rehearing en bane was granted, and we now affirm.
I.
A.
Background of the Parties and the Dispute. ORIX Credit Alliance is a subsidiary of companies that are in turn subsidiaries of ORIX Corporation, a Japanese company. App. 385-86, 616. ORIX Credit Alliance is a commercial finance company that is engaged primarily in the business of financing the acquisition or leasing of equipment. Id. at 310. ORIX Credit Alliance generally must boiTow the funds needed to support the financing it provides for its customers. Id. In simple terms, the company makes a profit by borrowing funds at one rate and then lending to its customers a higher rate. Id. at 80.
Frederick Keller was born on January 31, 1942. App. 610. After college, he was hired by Franklin National Bank and eventually handled its relationship with Credit Alliance Corporation, ORIX Credit Alliance’s predecessor. Id. at 611. In 1976, Credit Alliance Corporation hired Keller as a vice president, and in that capacity he shared the primary responsibility for raising funds for the company. Id. at 612, 1121. Keller obtained funding from banks, helped to supervise “the commercial paper program,” and worked on “other sources of funding.” Id. at 1120-21. The then-chairman of the company has described Keller’s work as “excellent,” and after several years, Keller was promoted to senior vice president of finance. Id. at 614.
In December 1984, First Interstate Ban-corp acquired Credit Alliance Corporation’s parent company, and Credit Alliance Corporation continued to do business as First Interstate Credit Alliance Corp. App. 613. After this acquisition, First Interstate Ban-corp provided most of the funding used by First Interstate Credit Alliance Corp. for its lending activities, and therefore it was no longer necessary for Keller to raise money. Id. at 614. Keller acquired the titles of executive vice president, chief financial officer, and chief credit officer, and he served as a member of the company’s board of directors. Id. at 70-71, 1121. In 1988, the chairman of the board of First Interstate Credit Alliance Corp. told Keller that he had been considered for the presidency of the company but that Daniel Ryan had been selected. Id. at 1121.
In September 1989, ORIX subsidiaries acquired First Interstate Credit Alliance Corp., which continued to do business under the name of ORIX Credit Alliance. App. at 616. Prior to the acquisition, Keller and six other key executives were requested to sign employment contracts with the new company, and Keller signed a three-year contract for employment at a substantial annual salary. Id. at 616, 644.
After this acquisition, Keller was given the responsibility for raising funding for ORIX Credit Alliance. App. 78. At that time, according to Keller’s affidavit, “Credit Alliance had approximately 1.6 billion in debt outstanding to First Interstate Bancorp, 1.3 billion of which was to continue to be provided on a temporary basis. Because it was the goal of Credit Alliance to obtain funding independent of First Interstate Bancorp and of ORIX Corp. or ORIX USA [an ORIX subsidiary] [Keller] determined that it would ultimately be necessary for Credit Alliance to have available credit facilities totaling approximately 1.5 billion dollars.” Id. at 616-17 (emphasis added). Keller stated that he communicated this goal to the board of directors at “[m]ore than one meeting.” Id. at 15.
Keller’s Failure to Meet the $1.5 Billion Goal and His Explanations. This goal, however, was never met or even approached. Keller himself stated in his deposition that “Credit Alliance never achieved the goal for funding that [he] had communicated to the board of- directors.” App. 16. Indeed, he acknowledged that, at all times from Septem[1104]*1104ber 1989 (the time of the ORIX acquisition) to April 1, 1993 (the time of his termination), funding provided by First Interstate Ban-corp and ORIX affiliates constituted more than 50% of ORIX Credit Alliance’s funding. Id. at 81-82. In December 1991, $785 million in credit facilities was available to ORIX Credit Alliance. App. 13. By September 1992 — approximately when the initial decision to terminate Keller was made — the total available bank lines had dropped to $695 million. Id. at 48. After September 1992, Keller did not secure any increase in bank lines. Id. at 24-25.
While Keller does not dispute that he failed to meet or approach the financing goal, he claims that this was due to factors beyond his control. See Appellant’s Br. at 8-11, 35-36; App. 617-32 (Keller Affidavit). For example, Keller explained that ORIX Credit Alliance was unable to launch a “commercial paper program,” as he had projected, because “there were many obstacles to obtaining a sufficiently high credit rating to permit Credit Alliance to issue and sell commercial paper on favorable terms.” App. 618. Among these, he stated, “were the absence of any guarantee by ORIX Corp., and the growing weakness of the Japanese economy which would affect Credit Alliance’s parent.” Id. Keller summarized these problems in memos that he sent to Ryan. Id.
Keller likewise provides a plethora of reasons for his failure to secure bank lines of credit. He cites the company’s credit rating, “the perceived ‘downturn in the equipment financing industry’ ... [,] Credit Alliance’s statistics for ‘past dues’ or untimely payments from its customers and other aspects of its portfolio ... [,] bank ‘environment[s] ... not conducive to risk of any sort’ ... [,] and bank limitations on lending to financing companies ... or to companies outside a particular geographical area,” “the negative impact of the recession in the United States and Japan during the late 1980’s and early 1990’s and the resulting reluctance of American banks to ‘book loans,’ ” and “banks’ reluctance to lend to a company having a Japanese parent, given the negative economic situation in Japan at the time.” App. 620-21.
During this same period, when Keller was allegedly unable to raise funds by means of a commercial paper program or bank lines of credit, Keller repeatedly expressed opposition to raising funds by “asset-backed securi-tization,” a process that involves the sale of accounts receivable or loan paper to a specially created trust that in turn sells interests or securities in that trust. See App. 90, 626. Ryan mentioned the possibility of raising funds in this way to Keller before or shortly after the ORIX acquisition {id. at 90), but Keller repeatedly advised Ryan that in his opinion asset-backed securitization was “not for us.” Id. at 28-29, 627.
Finally, Keller states that he explored the possibility of private placements of ORIX Credit Alliance debt with insurance companies and other institutional investors. App. 628. But Keller states that it was not until July 1992 (one or two months prior to the decision to terminate him) that he proposed to Ryan that ORIX Credit Alliance Corp. take “[t]he first step” in this direction, i.e., the selection of a bank to act as the company’s agent. Id. at 627-28.
ORIX Credit Alliance’s Assessment of Keller’s Performance. ORIX Credit Alliance points to evidence that paints a picture of growing dissatisfaction within the company about Keller’s failure to reach or approach the funding goal. Keller and Ryan both testified that Ryan repeatedly questioned Keller about the funding situation. App. 40, 44, 98-99. Ryan also stated that he asked Keller why ORIX Credit Alliance’s competitors were able to obtain forms of financing that his company either did not pursue or allegedly could not obtain. Id. at 90, 92, 98. See also id. at 627 (Keller Affidavit). One of ORIX Credit Alliance’s outside directors, David E. Mundell, who had served for nearly 20 years as the president of another leading commercial lending company, stated that at most, if not all, of the board meetings from March 1991 until April 1993 he questioned Keller and “expressed dissatisfaction with the lack of progress in raising funding.” Id. at 388. At one meeting, Mundell added, he “expressed the view that, based on [his] knowledge of the equipment finance industry and [his] experience in managing the liability side of finance companies’ balance sheets, [he] believed that Credit Alliance was not [1105]*1105raising funds in the amounts and on the terms that it should have been able to in light of the relevant factors, such as its financial statements, the sufficiency of its equity, and its status in the industry.” Id. Mundell continued that “Keller responded by offering a list of excuses for his inability to raise more funds,” but that he “did not ... make any concrete proposals for correcting the problems that, he claimed, were preventing him from producing the desired results.” Id. at 389.
Another outside director, Yoshiaki Ishida, who was also the president and chief executive officer of an ORIX parent corporation, stated that at several board meetings he “questioned Mr. Keller about his presentation.” App. 1164. Ishida added that he “was not satisfied with the results that Mr. Keller reported because the level of funds raised for Credit Alliance was much too low and the goal of independence in funding was not being achieved.” Id. Ishida added that he told Ryan at private meetings that “the ORIX parent companies were not happy with the lack of progress in raising funds for Credit Alliance, and with Credit Alliance’s continued reliance on another ORIX company (ORIX Ireland) for a large portion of its funding.” Id. Ishida also stated that he “told Mr. Ryan, on several occasions, that [he] felt that Mr. Keller’s work in raising funds was not satisfactory.” Id. at 1165.
In August or September of 1991, still another outside director, Sachio Hata, a senior officer of the parent Japanese corporation, suggested to Ryan that “perhaps [he] was remiss in giving Mr. Keller too much to do and that that could have been the reason why [the company’s] financing situation was making such little progress.” App. 103. Shortly after Hata made this remark, Ryan relieved Keller of his responsibilities as chief credit officer, primarily so that he “could focus on the financing function.” Id.
Keller attempts to counter this evidence by pointing to the absence of proof that Ryan ever expressly “criticized” his performance or disputed his explanations for his inability to obtain various types of funding. Keller points to Ryan’s inability during his deposition to recall more than two discussions at board meetings regarding Keller’s performance. App. 484-90, 492-93. In addition, Keller points out that, in describing those discussions, Ryan did not mention that either involved a direct criticism of Keller. See Appellant’s Br. at 12. Keller also notes that one ipember of the board of directors, Neil Umhafer, stated that during the period from 1988 to March 1992, “[n]o one challenged or disagreed with Keller’s presentations at [board] meetings or suggested in any way that the difficulties he was encountering were in any way due to his performance rather than factors beyond his control.”1 App. 1126.
Selection of the New Chief Operating Officer. In about December 1991 — in the midst of the time when Keller was experiencing difficulty in obtaining funding — Ryan announced that he planned to retire in approximately two and one-half years from his positions as chairman of the board and chief operating officer of ORIX Credit Alliance. App. 84. Ryan stated that he:
felt the job required someone that had a deep understanding and background of the company’s primary business, someone who had held a line position with the company, preferably someone who had personally performed as many of the tasks that are required to operate the company’s business as possible.
Id. at 101. He said that he therefore considered only the two most senior officers from the operational side of the company, division managers Philip Cooper, age 43, and Mark Lasher, age 50. Id. at 101, 268. Keller, then 50 years of age, was not considered even though he had been considered for the position of president in 1988 when Ryan was chosen. Id. at 101. Keller had never held a line position and had never worked in or managed any of the company’s branch offices or divisions. Id. at 65, 53.
Cooper was chosen as the new chief operating officer. App. 101. Cooper had decades of experience in line positions managing the company’s operations and had been with the company longer than Keller. Id. at 53, [1106]*1106101, 315. In May 1992, Keller, as a member of the board of directors, voted to ratify Cooper’s promotion. Id. at 314-15, 345.
The April IS, 1992 Conversation. Keller relies most heavily on a conversation he had with Ryan on April 13, 1992. Keller described this conversation as follows at his deposition:
[Ryan] assured me that he felt comfortable that Orix would be there for us as far as being able to loan us money. But then [he] made a comment that ... “We really can’t complain if we’re not out developing relationships.”
He said to me that he didn’t see me traveling around the country visiting with banks. He said I was spending a lot of time in New York City.
* * *
And then he said ... “If you are getting too old for the job, maybe you should go hire one or two young bankers.”
App. 27 (emphasis added).
Keller said that, because of Ryan’s reference to Keller’s age, Keller prepared a handwritten summary of the meeting within an hour or two after it ended. App. 25. These handwritten notes state in pertinent part:
DNR then suggested that we cannot complain about not being able to fund our needs if we have not made a good effort to develop lines etc. — he said I don’t see you traveling across ... country developing relationships I see you spending a lot of time in NYC. He suggested I hire one or two young bankers. Also discussed possibility of securitization if necessary.
Id. at 168-69. The handwritten summary contains no reference to Ryan’s alleged words “If you are getting too old,” but Keller explained at the deposition:
The reason I made those notes in the first place was because of that statement, I didn’t need these notes to remind me of what he said.
Id. at 26.
The Decision to Discharge Keller. By August or September 1992, Ryan had decided that Keller should be discharged. App. 91, 1156. After Ryan made his initial decision, he discussed it with at least four individuals who were directors or senior officers of the company, and all expressed agreement. Id. at 1156-57. Those informed included Mun-dell, Ishida, Cooper, and Jacob Mehl, an executive vice president and the general counsel of the company. Id.
Ryan and Cooper then drew up a list of criteria to be used in identifying a replacement. App. 316. Among their primary criteria were “experience in implementing asset-backed securitization programs and other creative forms of [fundraising],” “strong skills in working with rating agencies and bankers, particularly Japanese bankers,” and a “results driven” character. Id. at 316. According to Cooper, “[t]hese were among the areas in which we felt that Mr. Keller’s skills were inadequate for his job at Credit Alliance.” Id.
Cooper communicated these criteria to an executive search firm, which subsequently proposed several candidates. App. 316. From among these, Ryan selected Joseph McDevitt, who was born on December 12, 1946, and is thus four years, ten months, and 19 days younger than Keller. Id. at 317, 269, 610.
Keller’s Final Months. In September 1992 (i.e., at roughly the time when Ryan made his initial decision to terminate Keller), Ryan began to explore on his own the possibility of obtaining funding by means of asset-based securitization. App. 91. Shortly thereafter, Goldman Sachs & Co. was engaged by Cooper and Ryan for a securitization program, and within two years the company had closed two asset-backed secu-ritization deals and raised nearly $500 million. Id. at 389.
Keller was not informed of Ryan’s initial decision to discharge him, App. 590-92, and thus in late 1992 or early 1993 Keller presented to Cooper a one-page document entitled “TIMETABLE FOR DIVERSIFICATION OF FUNDING SOURCES.” This plan listed such items as the following:
2/8-5/1/93 Meet bankers and gauge level of interest in providing credit facilities.
2/9/93 Visit S & P; discuss 9-month results; determine feasibility of “A-2” rating before year-end numbers are available. If feasible, set up rating meeting [1107]*1107as soon as possible. If not feasible, see below.
3/15/93 Decide if public securitization is a viable funding source. If so, move to market (90 days).
App. 1107. In response to questioning by Keller’s attorney, Ryan acknowledged that most, if not all, of these steps were eventually taken, id. at 542-45, but Keller has not pointed to any evidence that any of the steps proposed were novel or that he did much, if anything, to accomplish any of the objectives.2
Keller was formally discharged by unanimous vote of the executive committee of the board of directors on April 1, 1993, effective on that date. App. 1159, 1162. Ryan went to Keller’s office to inform him of the decision. Id. at 591. Keller states that the following occurred:
I asked Dan if he was asking for my resignation because of my age and reminded him of the conversation that we had in April of ’92 when he was — when he asked me if I was getting too old for the job and suggested that I, if I were, that I hire one or two young bankers to travel around the country. He gave me — there was no response to that, to my comments.
Id. at 593.
Keller subsequently asked to “get together [with Ryan] to discuss [his] situation” (App. 595), and the two men met for lunch at a Manhattan restaurant on approximately April 9. Id. at 595-96. During the lunch, according to Keller, Ryan stated that Keller’s suggestion of a “$1 million plus” severance package was out of the question, and Ryan added:
“Look, you do what you have to do, you know. I have discussed this with Jerry Mehl, and he doesn’t see that we will have a problem.”
And then he said, “But, you know, Jerry is a lawyer and lawyers aren’t always right.”
Id. at 597.
McDevitt’s Performance. On April 5, 1993, McDevitt replaced Keller, and within a year he raised well over $1.5 billion in new credit facilities, including almost $2 million in bank lines, $250 million through a first offering of asset-backed securities, $275 million through a private placement of notes, $1 billion in syndicated credit facilities, and $300 million through the sale of commercial paper. App. 311-14.
B.
In August 1993, Keller commenced this action in federal district court, claiming that ORIX Credit Alliance denied him promotion to the position of chief operating officer and ultimately terminated him because of his age, in violation of the ADEA and the NJLAD. ORIX Credit Alliance moved for summary judgment, and the district court granted that motion.
With respect to Keller’s discharge claims, the district court first held that Keller had failed to make out a prima facie case under the scheme of proof set out in McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973). Specifically, the court held that Keller had not identified evidence showing that he was qualified for the position or that he was “ ‘replaced by someone significantly younger to permit an inference of age discrimination.’ ” Dist. Ct. Op. at 8 (quoting Gray v. York Newspapers, Inc., 957 F.2d 1070, 1088 (3d Cir.1992)). The court then concluded that, even if Keller had established a prima facie case, ORIX Credit Alliance had proffered a legitimate business reason for Keller’s dismissal, namely, “Keller’s failure to make adequate progress toward the $1.5 billion independent financing goal,” and that Keller had not pointed to evidence that a reasonable jury could view as establishing pretext or as proving that his discharge was due to age discrimination. Id. at 10-11.
With respect to Keller’s denial-of-promotion claim, the court reasoned that the same evidence of Keller’s failure to meet or approach the financing goal was sufficient to warrant summary judgment on that claim as well. Id. at 11. Keller then took this appeal.
[1108]*1108II.
We turn first to Keller’s discharge claim. Keller contends that this claim should have survived summary judgment under either McDonnell Douglas or Price Waterhouse v. Hopkins, 490 U.S. 228, 109 S.Ct. 1775, 104 L.Ed.2d 268 (1989).
A. McDonnell Douglas
In McDonnell Douglas, the Supreme Court created a special scheme for structuring the presentation of evidence in discriminatory treatment cases under Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e-1 et seq. Our court has applied a slightly modified version of this scheme in ADEA cases. See, e.g., Waldron v. SL Industries Inc., 56 F.3d 491, 494-95 (3d Cir.1995); Sempier v. Johnson & Higgins, 45 F.3d 724, 728 (3d Cir.), cert. denied, 515 U.S. 1159, 115 S.Ct. 2611, 132 L.Ed.2d 854 (1995); Torre v. Casio, Inc., 42 F.3d 825, 829-30 (3d Cir.1994); Healy v. New York Life Ins. Co., 860 F.2d 1209, 1214 (3d Cir.1988).3 Cf. O’Connor v. Consolidated Coin Caterers Corp., 517 U.S. 308, -, 116 S.Ct. 1307, 1310, 134 L.Ed.2d 433 (1996) (assuming arguendo that McDonnell Douglas applies under ADEA).
The McDonnell Douglas scheme has three steps. First, the plaintiff must produce evidence that is sufficient to convince a reasonable factfinder to find all of the elements of a prima facie case. St. Mary’s Honor Center v. Hicks, 509 U.S. 502, 506, 113 S.Ct. 2742, 2746-47, 125 L.Ed.2d 407 (1993). When the plaintiff alleges unlawful discharge based on age, the prima facie case requires proof that (i) the plaintiff was a member of the protected class, i.e., was 40 years of age or older (see 29 U.S.C. § 631(a)), (ii) that the plaintiff was discharged, (iii) that the plaintiff was qualified for the job, and (iv) that the plaintiff was replaced by a sufficiently younger person to create an inference of age discrimination. Sempier, 45 F.3d at 728.
If the plaintiff offers sufficient proof of these elements, step two is reached. The burden of production (but not the burden of persuasion) shifts to the defendant, who must then offer evidence that is sufficient, if believed, to support a finding that it had a legitimate, nondiscriminatory reason for the discharge. Hicks, 509 U.S. at 506-07, 113 S.Ct. at 2746-47. If the defendant cannot satisfy this burden, judgment must be entered for the plaintiff. Id. at 509, 113 S.Ct. at 2748. On the other hand, if the defendant does satisfy this burden, step three is reached. The plaintiff may then survive summary judgment or judgment as a matter of law by submitting evidence
from which a factfinder could reasonably either (1) disbelieve the employer’s articulated legitimate reasons; or (2) believe that an invidious discriminatory reason was more likely than not a motivating or determinative cause of the employer’s action.
Fuentes v. Perskie, 32 F.3d 759, 763 (3d Cir.1994). Accord Sheridan v. E.I. DuPont de Nemours and Co., 100 F.3d 1061, 1067 (3d Cir.1996) (en banc), cert. denied, - U.S. -, 117 S.Ct. 2532, 138 L.Ed.2d 1031 (1997).
In this appeal, we find it unnecessary to consider steps one and two of the McDonnell Douglas scheme. Step two is not contested, and although the parties dispute whether Keller met step one, we will assume for the sake of argument that he did, because we agree with the district court that Keller did not satisfy step three under either the first or second prong of the Fuentes test.
1. Prong One. As noted, a plaintiff may satisfy this prong by offering evidence “from which a factfinder could reasonably ... disbelieve the employer’s articulated legitimate reasons.” Fuentes, 32 F.3d at 764. But as we have explained:
To discredit the employer’s proffered reason ... the plaintiff cannot simply show that the employer’s decision was wrong or mistaken, since the factual dispute at issue is whether discriminatory animus motivated the employer, not whether the employer is wise, shrewd, prudent, or competent. Rather, the nonmoving plaintiff must demonstrate such weaknesses, implausibilities, inconsistencies, incoherencies, or eontra-[1109]*1109dictions in the employer’s proffered legitimate reasons for its actions that a reasonable factfinder could rationally find them unworthy of credence.
Id. at 765. As another court of appeals has put it, “federal courts are not arbitral boards ruling on the strength of ‘cause’ for discharge. The question is not whether the employer made the best, or even a sound, business decision; it is whether the real reason is [discrimination].” Carson v. Bethlehem Steel Corp., 82 F.3d 157, 159 (7th Cir.1996).
The defendant in this case provided evidence that it had a particularly powerful reason for discharging Keller, i.e., his failure to meet or even approach the goal of raising $1.5 billion in financing. As previously noted, ORIX Credit Alliance makes a profit by borrowing money and then lending it at a higher rate. Consequently, borrowed money is the company’s life blood, and it thus seems clear (and we do not understand Keller to disagree) that the company would have had a strong reason for discharging a key executive who unjustifiably failed to meet a reasonable objective relating to the raising of funds. Moreover, Keller cannot argue that the objective of raising $1.5 billion was unreasonable when it was originally set: he himself set that goal, and he assured the board of directors that he could meet it.
Instead, Keller makes two chief arguments: first, that the evidence in the summary judgment record shows that his inability to meet or approach the $1.5 billion objective was due to factors beyond his control and, second, that evidence in the summary judgment record shows that Ryan knew that this was so. We will discuss each of these arguments in turn.
Evidence that Keller’s failure to reach or approach the $1.5 million goal was due to factors beyond his control. In considering this argument, it is critical to keep in mind that the question under the first prong of the Fuentes test is not whether “the employer’s decision was wrong or mistaken.” Fuentes, 32 F.3d at 765. Accordingly, Keller cannot survive summary judgment under this prong simply by pointing to evidence that could convince a reasonable factfinder that he did as well as he could under the circumstances. Rather, he “must demonstrate such weaknesses, implausibilities, inconsistencies, incoherencies, or contradictions in the employer’s proffered legitimate reasons for its action that a reasonable factfinder could rationally find them ‘unworthy of credence.’ ” Fuentes, 32 F.3d at 765. In simpler terms, he must show, not merely that the employer’s proffered reason was wrong, but that it was so plainly wrong that it cannot have been the employer’s real reason.
When this point is kept in mind, it is apparent that Keller failed to satisfy prong one of Fuentes. Whether Keller could have met or come close to the $1.5 billion goal under the business conditions that prevailed from 1989 to 1992 is a complicated question that would be difficult to resolve without expert testimony of a sort that is lacking in the summary judgment record of this case. But the relevant question is not whether Keller could have done better; instead, the relevant question is whether the evidence shows that it was so clear that Keller could not have done better that ORIX Credit Alliance could not have believed otherwise. The answer to this question is plainly negative.
The evidence relating to asset-backed sec-uritization illustrates the weakness of Keller’s position. It is undisputed that Ryan had an early interest in this method of raising funds. Ryan discussed asset-backed sec-uritization with Keller before or shortly after the ORIX acquisition, and it is also undisputed that Keller rejected this idea. Ryan testified without contradiction that on many subsequent occasions he raised the possibility of asset-backed securitization with Keller. Ryan said that every time he read an article in the Wall Street Journal about a competitor’s utilization of this technique, he mentioned the subject to Keller, App. 90, and it is undisputed that Keller’s consistent response was that asset-backed securitization should not be pursued by ORIX Credit Alliance, in part because of the nature of its business. Id. at 28. Indeed, Keller’s own affidavit reiterates this position. Id. at 626-28. Finally, in September 1992, Ryan decided to explore the matter on his own without Keller’s knowledge or participation. Ryan met with leading investment banking firms, en[1110]*1110gaged the services of one such firm, and within two years the company raised nearly $500 million through two offerings of asset-backed securities. In the face of this evidence, a reasonable factfinder could not find that ORIX Credit Alliance’s dissatisfaction with Keller’s failure to pursue asset-backed securitization was so clearly wrong that it cannot have been sincere.
Keller’s evidence relating to bank lines of credit likewise falls short of what would be necessary to show that ORIX Credit Alliance’s dissatisfaction with his performance was so clearly unfounded that it cannot have been sincere. Keller’s brief states:
[BJanks told Keller that they were not interested in companies outside their region; that their “credit culture [had] bec[o]me very conservative;” that they were troubled by the Japanese economy and the level of delinquent accounts at Credit Alliance, and sought business only with highly rated companies; that they perceived a “downturn in the equipment financing industry ... which they expect will become even worse.”
Appellant’s Br. at 10 (footnotes omitted).
This recital is based on four file memos written by Keller during a period of more than three years. See Appellant’s Br. at 10 & nn. 7-10. The first memo relates that an officer of a regional bank with offices in Florida and Georgia told Keller that his bank “only does business with local companies or national companies with local (Florida/Georgia) operations.” App. 394. The second memo did not report that the bank in question had refused to extend credit; instead, it concluded by saying: “We decided to meet again after [the bank officer] has received and reviewed our 1992 financial information.” Id. at 401. The third memo, dated several months after the decision to terminate Keller was made, does recount that an officer at a major bank “repeated his many stories as to why it [was] difficult for him to get a credit facility approved for our company.” Id. at 402. The final memo stated that an officer at a major bank told Keller (in 1990) that “the Bank would prefer to delay providing ... a line of credit.” Id. at 666. Taken together, these file memos and the additional memos cited in Keller’s affidavit (see id. at 620-21) constitute evidence from which a reasonable factfinder could conclude that Keller experienced difficulties in securing bank lines, but they could not persuade a reasonable factfin-der that it was so plainly impossible for Keller to secure additional lines of credit from other banks during the 1989-1992 period that ORIX Credit Alliance’s dissatisfaction with his performance must not have been real.
We will not discuss the evidence relating to Keller’s failure to obtain more funding by other means. However, after examining all of the evidence identified in Keller’s brief for the purpose of showing that his failure to meet or approach the $1.5 billion financing goal was due to factors beyond his control, we are convinced that Keller has not shown that it was so plain that he could not have done substantially better under the circumstances that ORIX Credit Alliance could not have truly believed otherwise.
Evidence that Ryan knew Keller could not have done better. Keller also argues in his brief that the summary judgment record contains evidence that Ryan knew that Keller could not have done appreciably better under the business circumstances that prevailed from 1989 through 1992. See Appellant’s Br. at 8-9. This argument, however, is simply not supported by the record.
Keller’s brief states that “Ryan repeatedly acknowledged” that “Keller’s inability to obtain more credit on favorable terms was due to circumstances beyond his control.” Appellant’s Br. at 35-36. But our review of the record citations provided in Keller’s brief has not disclosed a single such acknowledgment. Instead, most of the record citations are based on passages from Ryan’s deposition during which the following occurred. Keller’s attorney showed Ryan documents that had been written by Keller and that memorialized statements that had allegedly been made by third parties, such as officers of banks or rating agencies, and that explained why these third parties were unwilling to take various actions that would have been favorable to ORIX Credit Alliance. Keller’s attorney then asked Ryan whether he had any basis for disputing the accuracy of the documents, and Ryan (who generally had no recollection of previously seeing the docu[1111]*1111ments) said that he had no basis for disputing their accuracy. See Appellant’s Br. at 10 & nn. 7-10 (citing App. 554-55, 558-59, 561-62, 563-646).
These exchanges merely show that Ryan did not dispute the accuracy of particular documents that recounted a limited number of specific statements allegedly made over the course of several years by individuals associated with particular banks and rating agencies. A factfinder could not reasonably draw from these exchanges the general conclusion that Ryan “repeatedly acknowledged” that “Keller’s inability to obtain credit on favorable terms was due to circumstances beyond his control.” Appellant’s Br. at 35-36.
Keller also relies on the assertion that his performance was never “criticized” prior to the April 13 meeting with Ryan, and he argues that this absence of criticism shows that ORIX Credit Alliance was not sincerely disturbed by his failure to approach the $1.5 billion goal. Appellant’s Br. at 29, 36. The summary judgment record contains evidence that Keller was criticized (director Mundell’s comments are perhaps the clearest example), but Keller claims that he was never “criticized,” except at the April 13, 1992 meeting, and in the present procedural posture of the case, we accept Keller’s position.
Keller does not dispute, however, that he was repeatedly “questioned” by Ryan and others about this matter, and therefore the question is to what extent a reasonable fact-finder could infer from the absence of criticism (as distinct from questioning) that ORIX Credit Alliance was not really troubled by Keller’s failure to approach the $1.5 billion goal. We conclude that a reasonable factfinder could draw only a relatively weak inference. Employers who are dissatisfied with the performance of their employees sometimes voice express criticism to those employees, but employers do not always do so. See Healy, 860 F.2d at 1216 (“The company is under no obligation to warn plaintiff of complaints regarding his performance and, if anything, the effect of such evidence is equivocal, perhaps indicating that plaintiff was receiving the benefit of the doubt.”) (citation omitted). Evidence that a plaintiff was not criticized may take on significance if the plaintiff can show that other comparable employees regularly received express evaluations of their work, but Keller does not point to any such evidence. Moreover, in light of the patent importance of the $1.5 billion goal, and in light of the steady “questioning” of Keller about this matter, the absence of explicit criticism cannot reasonably be viewed as having great importance.
In sum, after considering all of the evidence that has been called to our attention, we conclude that a reasonable factfinder could not find that the words, actions, or omissions of the relevant ORIX Credit Alliance officers evidenced their belief that Keller was doing as well as could be expected under the circumstances. For this reason and the others explained above, we therefore hold that Keller cannot defeat summary judgment based on the first prong of the Fuentes test.
2. Prong Two. Accordingly, we proceed to the question whether Keller can survive summary judgment under prong two of the Fuentes test. Under this prong, Keller must identify evidence in the summary judgment record that “allows the fact finder to infer that discrimination was more likely than not a motivating or determinative cause of the adverse employment action.” Fuentes, 32 F.3d at 762. In other words, under this prong, Keller must point to evidence that proves age discrimination in the same way that critical facts are generally proved — based solely on the natural probative force of the evidence.
Keller’s best evidence under prong two is his account of his conversation with Ryan on April 13, 1992. As previously noted, Keller testified that Ryan made the following comments:
“We really can’t complain if we’re not out developing relationships.”
He said to me that he didn’t see me traveling around the country visiting with banks. He said I was spending a lot of time in New York City____
And then he said ... “If you are getting too old for the job, maybe you should hire one or two young bankers.”
App. 27 (emphasis added). Although Ryan denied using the words “If you are getting [1112]*1112too old for the job,” and although Keller’s contemporaneous notes of the conversation omit any mention of this phrase, we are required, in reviewing the district court’s grant of summary judgment in favor of ORIX Credit Alliance, to accept Keller’s account of the conversation.
Ryan’s alleged words certainly constitute evidence from which a reasonable factfinder could draw an inference of age-based animus, but we do not think that these words alone could reasonably be viewed as sufficient to prove by a preponderance of the evidence that age was a determinative cause of Keller’s subsequent termination. For one thing, the alleged comment occurred four or five months prior to the time when Ryan decided that Keller should be discharged. In addition, the alleged remark did not refer to the question whether Keller should be retained or fired but instead concerned the hiring of other employees to assist him. Furthermore, the alleged statement pertained to only one method of raising funds — obtaining lines of credit from banks outside New York City by traveling to meet their officers. Even if Ryan’s alleged statement is interpreted to mean that he felt that Keller might be getting too old to do the traveling necessary to raise funds in this way, no evidence has been brought to our attention that other methods of raising funds, such as beginning a commercial paper program or utilizing asset-backed securitization, would have required extensive travel.
Keller’s remaining evidence under prong two is insubstantial. Keller’s statistical evidence is of little if any value.4 Moreover, we reject Keller’s suggestion that Ryan’s actions during his meeting with Keller on April 1, 1993, when he asked for Keller’s resignation, and his comments during their subsequent restaurant meeting approximately one week later constitute significant evidence of age-based animus.
During the meeting on April 1, 1993, Ryan gave Keller a draft letter of resignation to consider. See App. at 594. Under the terms set out in this letter, Keller would have received certain substantial benefits, including one-half his annual salary plus $50,000. See App. 1114. In return, Keller would have released ORIX Credit Alliance from all claims. Id. at 1115. Noting that this blanket release would have presumably included claims of age discrimination, Keller seems to imply that the inclusion of this provision in the letter evidences ORIX Credit Alliance’s awareness that Keller had grounds for an age-discrimination claim against it. Appellant’s Br. at 4. This implication is far-fetched. Without evidence that a request for a blanket release is not a common practice when an executive is asked to resign under terms such as those set out in the letter, the inclusion of this clause in the proposed letter of resignation has little evidentiary worth.
Likewise, we reject Keller’s argument that Ryan’s comments during the restaurant meeting evidenced consciousness of guilt of age discrimination. See Appellant’s Br. at 4, 35. At the April 1, 1993 meeting, Keller had asked Ryan:
“if he was asking for [Keller’s] resignation because of [his] age and [Keller] reminded him of the conversation that [they] had in April of ’92 when ... he asked [Keller] if [he] was getting too old for the job and suggested that ... if [he] were, that [he] hire one or two young bankers.”
Id. at 593. At his deposition, Ryan stated that he assumed, based on Keller’s comment, that Keller “was thinking of an age discrimination suit.” Id. at 512-13. At the subsequent restaurant meeting, after rejecting Keller’s request for a “$1 million plus” severance package, Ryan said:
Look, you do what you have to do ... I have discussed this with Jerry Mehl [the ORIX Credit Alliance general counsel] and he doesn’t see that we will have a problem • • • [b]ut, you know, Jerry is a lawyer and lawyers aren’t always right.
Referring to these events, Keller’s brief states:
Ryan’s comments that Credit Alliance could be found liable for age discrimination [1113]*1113are evidence that, at a minimum, he indeed had made the biased statement Keller attributed to him.
Appellant’s Br. at 4. Keller further argues that evidence of age discrimination is provided by “Ryan’s statements at the time of Keller’s discharge that he assumed that Keller would sue for age discrimination.” Appellant’s Br. at 35 (footnote omitted). These arguments have no merit.
When Keller asked at the April 1, 1992 meeting whether he was being fired because of his age, any reasonable person would have realized that Keller might thereafter sue for age discrimination. Thus, Ryan’s assumption that Keller might file such a suit hardly constitutes evidence of consciousness of guilt.
Furthermore, Ryan’s statement that his company’s general counsel might turn out to be wrong in predicting that Keller’s termination would not cause a “problem” has little if any evidentiary value to show that Ryan believed that Keller had a meritorious age-discrimination claim. Needless to say, even an ultimately unsuccessful claim may constitute a “problem,” and due to the vagaries of the legal process, unmeritorious suits are not always unsuccessful (just as meritorious suits do not always succeed).
In assessing whether the proof in this case is sufficient to establish by a preponderance of the evidence that age was a determinative cause of Keller’s termination, a reasonable factfinder would have to consider, in addition to the evidence noted above, the proof underlying the elements of the prima facie case. Thus, a reasonable factfinder would have to weigh the fact that Keller, who was 51 years old when fired, was replaced by a man who was about four years and ten and one-half months younger.
Finally, a reasonable factfinder would also have to consider the evidence, which we discussed in part IIA1 of this opinion, that ORIX Credit Alliance had a powerful, legitimate reason for discharging Keller, namely, his failure to meet or even approach the critical $1.5 billion goal that he himself had set. A reasonable factfinder would have to ask whether a company like ORIX Credit Alliance was more likely to be concerned about Keller’s failure to raise these funds or about replacing him with a man who was some four years and ten and one-half months younger.
Considering all of the evidence that is relevant with respect to prong two, we conclude that a reasonable factfinder could not find that the proof is sufficient to establish by a preponderance of the evidence that age was a determinative factor in Keller’s termination. Consequently, we hold that Keller cannot survive summary judgment under prong two of the Fuentes test. Since we have already held that he failed under prong one as well, it follows that he cannot defeat summary judgment under the scheme of proof set out in McDonnell Douglas.
B. Price Waterhouse
We therefore move on to Keller’s argument that he was entitled to survive summary judgment under Price Waterhouse. Under Justice O’Connor’s controlling opinion in Price Waterhouse, if a plaintiff “show[s] by direct evidence that an illegitimate criterion was a substantial factor in the decision,” the burden of persuasion shifts to the employer “to show that the decision would have been the same absent discrimination.” 490 U.S. at 276, 109 S.Ct. at 1804 (O’Connor, J. concurring) (emphasis added). See Armbruster v. Unisys Corp., 32 F.3d 768, 778 (3d Cir.1994). The precise meaning of Justice O’Connor’s term “direct evidence” has divided the courts. See Linda Hamilton Krieger, The Content of Our Categories: A Cognitive Bias Approach to Discrimination and Equal Employment Opportunity, 47 Stan. L.Rev. 1161, 1220-21 (1995) (describing the varying approaches of the circuits); Note, Despite the Smoke, There Is No Gun: Direct Evidence Requirements in Mixed-Motives Employment Law After Price Waterhouse v. Hopkins, 46 Stan. L.Rev. 959, 970-79 (1994) (same). Similarly, when the present ease was before the panel, the majority and the dissent disagreed on the question whether Ryan’s alleged statement on April 13, 1992, constituted “direct evidence” within the meaning of Price Waterhouse.
On reconsidering this case en banc, we conclude that it is not necessary for us to resolve this question. We have held in part IIA2 of this opinion that a reasonable jury [1114]*1114could not find by a preponderance that age was a determinative factor. If we held that Keller provided “direct evidence” within the meaning of Price Waterhouse, Keller could avoid summary judgment only if a reasonable jury could fail to find by a preponderance that age was not a determinative factor. Here, for the reasons explained above in part IIA2 of this opinion, a reasonable factfinder could not fail to find by a preponderance that age was not a determinative factor in Keller’s termination. We therefore hold that Keller cannot survive summary judgment on his discharge claim under Price Waterhouse.
III.
We proceed finally to Keller’s claim that ORIX Credit Alliance failed to promote him to the position of chief operating officer in May 1992 because of his age. Assuming for the sake of argument that Keller could make out the elements of a prima facie ease with respect to this promotion decision, we hold that ORIX Credit Alliance proffered a legitimate explanation for the decision and that Keller did not satisfy either prong one or two of the Fuentes test.5
Ryan explained that he felt that the job of chief operating officer “required someone [who] had a deep understanding and background of the company’s primary business, someone who held a line position with the company, preferably someone who had personally performed as many of the tasks that are required to operate the company’s business as possible.” App. 101. Keller has not pointed to any evidence showing that ORIX Credit Alliance did not in fact rely on this criteria in choosing the new chief operating officer. Nor has Keller pointed to any evidence that he possessed such experience. Furthermore, the selection of the new chief operating officer came at a time when Keller was failing in the performance of the job he then held. Months earlier, he had been relieved of his responsibilities as chief credit officer so that he could focus on raising funds, and by May 1992, it is undisputed that' Keller was being repeatedly questioned about his failure to meet or approach the $1.5 billion target. Under these circumstances, it is apparent that the company had legitimate reasons for failing to promote Keller to the top position of chief operating officer. Thus, Keller failed to satisfy prong one of the Fuentes test.
We likewise hold that Keller failed to meet prong two of that test. We have already discussed all of the evidence on which Keller relies to show age discrimination, and we will therefore not discuss that evidence again here. Considering all of that evidence, and keeping in mind that Ryan’s alleged comment on April 13, 1992 came only a few weeks before the promotion decision was made, we nevertheless conclude that the evidence is insufficient to convince a reasonable factfinder by a preponderance that age was a determinative factor in the promotion decision.
IV.
For the reasons explained above, we therefore affirm the decision of the district court granting summary judgment in favor of ORIX Credit Alliance on all of Keller’s claims.