On April 9, 1982, the plaintiff, Irving Gray, a Right-of-Way Agent in New Hampshire for the defendant, New England Telephone & Telegraph (hereinafter, “NET”), was discharged from his employment. Gray had been employed continuously since 1955, and was 51 years of age at the time of his dismissal.
On September 23, 1982, Gray filed suit on three theories: (1) breach of employment contract; (2) violation of the Age Discrimination in Employment Act (hereinafter, “ADEA”), 29 U.S.C. §§ 621-634; and (3) negligent or malicious refusal to provide notice of pension benefits in violation of the Employment Retirement Income Security Act (hereinafter, “ERISA”), 29 U.S.C. §§ 1001-1461. Gray was originally represented by counsel, but dismissed his attorneys in the fall of 1984. He was then given time to obtain alternate counsel, and the trial, scheduled for December, was postponed until July 1985. At the final pretrial conference held in June 1985, Gray advised the court that he had yet to obtain alternate counsel. The district court judge indicated that the plaintiff would have to proceed on the scheduled trial date, even if he must proceed pro se.
At the trial, Gray appeared pro se and presented several witnesses and a number of exhibits. At the close of the plaintiff's case, the district court judge directed a verdict against Gray on his claims of wrongful discharge and ERISA violations, but found that Gray had established a prima facie case as to the ADEA claim. NET then presented its defense of the ADEA claim, asserting a legitimate nondiscriminatory reason for Gray’s discharge: that Gray had violated a certain “back-to-work” agreement and various other company policies. At the close of all the evidence, the district court judge directed a verdict against Gray on the ADEA claim, finding *253that there was no evidence upon which a reasonable jury could determine that the discharge of Gray occurred in violation of the ADEA.
On this appeal, three issues are raised. Gray claims that the district court judge erred in directing a verdict at the close of the evidence on the ADEA claim. He also claims that the court erred in requiring Gray to proceed to trial without counsel. Finally, NET claims error in the district court’s failure to award it attorney’s fees. We affirm the district court in all respects.
The Directed Verdict
The standard of review of a trial court’s grant of a directed verdict is well established in this circuit and was most recently articulated in Goldstein v. Kelleher, 728 F.2d 32, 39 (1st Cir.1984) (quoting Carlson v. American Safety Equipment Corp., 528 F.2d 384, 386 (1st Cir.1976) (citations omitted), cert. denied, — U.S.-, 105 S.Ct. 172, 83 L.Ed.2d 107 (1984):
In order to uphold a grant of a directed verdict, we must find that, viewing the evidence in the light most favorable to the non-moving party, reasonable jurors could come to but one conclusion. We must give plaintiff the benefit of every legitimate inference. However, such inferences may not rest on conjecture or speculation, but rather the evidence offered must make “the existence of the fact to be inferred more probable than its nonexistence.”
Additionally, in reviewing the evidence, the trial judge must come to his or her determination without evaluating the credibility of the witnesses or considering the weight of the evidence. deMars v. Equitable Life Assurance Society, 610 F.2d 55, 57 (1st Cir.1979).
Keeping these principles in mind, our review of the record shows that the plaintiff put forward the following evidence bearing on his ADEA claim. NET was aware of Gray’s age — 51. Gray had a history of evaluations dating from 1975 to 1980 characterizing his job performance as “completely satisfactory” and which at various times commended him for his attendanee record, his safety record, and his diligence in securing easements. There was also testimony that after Gray was dismissed, his job was still extant and was filled by two contract workers aged 58 and 65, who were not salaried employees and did not receive regular benefits.
Gray also introduced the testimony of Paul Donaldson, a senior marketing manager for the firm, Prime Computer. In July 1981, Donaldson was contacted by William Adams, a security manager with NET who was investigating Gray’s use of NET phone privileges. Adams contacted Donaldson to determine whether certain calls made by Gray to Donaldson had been business or personal calls. Donaldson testified that, in the course of this telephone interview, Adams
indicated to [Donaldson] that it was his perception, and I assume the phone company’s, because he was speaking for them, that Bud Gray had been a valued employee at one point in time, and had a good record, but had literally outlived his usefulness, that it was time to make way for new people within the ranks____
Joint Appendix (hereinafter, “J.A.”) at 129. Donaldson further testified as to Adams’ statements about Gray’s inability to get along with his supervisor because of age differentials.
[I]t was [Adams’] perception, and again I assume the phone company’s, that there was a problem between [Gray and his supervisor], because either she had a difficult time relating to [Gray] because [Gray] [was] old enough to be her father, as he stated, or [Gray] to [his supervisor] because she was young enough to be [Gray’s] daughter, and [Adams] felt that the age situation probably was an issue in ... his stated inability for [Gray] to either get along or them to get along with [Gray].
J.A. at 132. Gray himself testified along the same lines asserting that he distinctly remembered his supervisor expressing displeasure over Gray’s salary being greater than her own.
*254There was also evidence in the record about a certain “Management Income Protection Plan” (hereinafter, “MIPP”) offered by NET. This program, according to Norman Clark, a first level management supervisor, was developed to encourage people to retire early on a voluntary basis so as to reduce the work force; it was directed at employees in the 50 plus age range. Wilfred Ott, a retired network manager of NET, pointed to MIPP as an indication of company problems with age discrimination: “the company took steps to ease people out on retirement, which to me involved age. We didn’t want to spend the money to retrain them and go on.” J.A. at 67. Ott admitted that he had no expertise with respect to MIPP.
Finally, Gray presented testimony refuting each of the reasons NET asserted as legitimate, non-discriminatory bases for discharging Gray. He claimed that he only did noncompany work, i.e., the serving of process as a deputy sheriff for Hillsborough County, during his lunch hour, rather than on company time. He claimed to be legitimately spending his time at the Hills-borough County court house in legal research related to his job, rather than engaging in activities unrelated to his work for NET. He presented evidence that the company vehicle was in such poor condition that Gray was justified in refusing to drive it. And he also elicited testimony that personal errands which he performed during working hours were customarily done by other NET employees during those hours.
In determining whether the directed verdict can be upheld, we must next ascertain whether the facts presented by the plaintiff sufficiently satisfied his burdens of proof under the ADEA so that the case should have gone to the jury. The district court held that Gray had presented sufficient evidence to establish a prima facie case of discrimination. It also found that NET had met its burden of going forward with evidence to rebut the inference of discrimination, by articulating legitimate, nondiscriminatory reasons for its discharge of Gray. Thus, the burden of production shifted back to Gray to show that the legitimate reason proferred by NET was not the true reason for the discharge decision, but, rather, was a mere pretext, and that Gray’s age was a determinative factor in NET’s employment decision. Gray, as the plaintiff, of course, retained the ultimate burden of persuasion throughout the case. See Loeb v. Textron, Inc., 600 F.2d 1003 (1st Cir. 1979).
Because the prima facie case established by Gray was rebutted by the defendant, the initial inference or presumption of discrimination dissolved. See Furnco Construction Corp. v. Waters, 438 U.S. 567, 577, 98 S.Ct. 2943, 2949, 57 L.Ed.2d 957 (1978); Smithers v. Bailar, 629 F.2d 892 (3d Cir.1980). And therefore, the establishment of a prima facie case could not be considered the equivalent of an ultimate finding of discrimination. Furnco, 438 U.S. at 577, 98 S.Ct. at 2949. Consequently, whether the directed verdict was properly given must be evaluated by looking at whether the plaintiff presented sufficient evidence as to his ultimate burden of proof on this claim, rather than merely as to his prima facie case. Cf. Dance v. Ripley, 776 F.2d 370, 373 (1st Cir.1985) (Title VII case emphasizing importance of focusing on ultimate burden of proof). The evidence presented as part of the plaintiff’s prima facie case, however, may still be considered in determining whether Gray presented sufficient evidence to go to the jury on his ultimate burden; its inherent probative force is not automatically destroyed with the destruction of the original presumption of discrimination. See Texas Department of Community Affairs v. Burdine, 450 U.S. 248, 255 n. 10, 101 S.Ct. 1089, 1095 n. 10, 67 L.Ed.2d 207 (1981).
Considering the evidence presented by Gray and the applicable law relating to ADEA claims, we agree with the district court judge that a directed verdict was appropriate in this ease. The only direct evidence of age discrimination was the Donaldson testimony. Assuming the truth of the facts to which Donaldson testified along with all reasonable inferences there*255from, as we are bound to do in reviewing a directed verdict, Goldstein v. Kelleher, 728 F.2d at 39, the most which could be concluded is that Adams, a management security officer at NET, felt Gray had outlived his usefulness and had difficulties dealing with his supervisor because of age differences. There is nothing in the record to indicate that the company or those responsible for discharging Gray had these same beliefs or that they communicated them to Adams. Donaldson specifically prefaced his remarks to the effect that “it was [Adams’] perception”; Donaldson only assumed that these perceptions were also those of the company. A reasonable inference cannot be drawn from these isolated statements (made by a lower level management employee in a department which had nothing to do with the discharge decision many months before Gray was fired) that age discrimination was a determinative factor in NET’s decision to fire Gray. Also, the statements about Gray’s age difficulties with his supervisor suggest that it could equally have been discrimination on the part of Gray rather than discrimination by the supervisor which resulted in the friction.
In addition, no permissible inference could be drawn from the evidence about MIPP. The record shows only that MIPP was a voluntary, early-retirement program. The record is devoid of evidence that the program was used by NET to force older employees out of the company or that this program was used to force Gray out of the company in this particular case. Wilfred Ott’s testimony about his own feelings about MIPP when he had no knowledge or expertise about its workings or effects is, standing alone, simply not enough to permit an inference of age discrimination. Absent evidence of illegal discrimination, the ADEA does not prohibit the formulation of such voluntary plans.
The plaintiff suggests, further, that discrimination could be inferred from the fact that Gray’s replacements were contract, rather than salary, employees, and received no pension and fringe benefits. Therefore, Gray argues, NET saved money by discharging him, and such a decision constitutes illegal age discrimination. In making this argument, Gray relies on Dace v. ACF Industries, Inc., 722 F.2d 374 (8th Cir. 1983). In that case the Eighth Circuit reversed the grant of a directed verdict because the plaintiff had presented specific evidence that the company demoted the plaintiff to save money and linked that evidence to the discrimination alleged. Because of the existence of this evidence, the Eighth Circuit ruled that a jury could have found that there indeed had been a violation of the ADEA.
In this case, however, there is no evidence in the record indicating that NET would save money by demoting Gray and replacing him with contract employees. Nor was a link shown between the savings NET would achieve by firing Gray and Gray’s actual discharge. The plaintiff cannot make inferences from facts which do not appear in the record.
Finally, although Gray did present evidence contesting the factual underpinnings of the reasons proferred by NET for firing Gray, these, without more are insufficient in this case to present a jury question. As stated in Loeb v. Textron, 600 F.2d 1003, 1012 n. 6 (1st Cir.1979), “[w]hile an employer’s judgment or course of action may seem poor or erroneous to outsiders, the relevant question is simply whether the given reason was a pretext for discrimination.” It is not enough for the plaintiff to show that the employer made an unwise business decision, or an unnecessary personnel move. Nor is it enough to show that the employer acted arbitrarily or with ill will. These facts, even if demonstrated, do not necessarily show that age was a motivating factor. See Chipollini v. Spencer Gifts, 613 F.Supp. 1156 (D.N.J. 1985); cf. Kephart v. Institute of Gas Technology, 630 F.2d 1217, 1223 (7th Cir. 1980) (in context of discussion of prima facie ADEA claim, noting that ADEA not intended as vehicle for judicial review of business decisions and that material issue of fact is not raised merely by challenging *256judgment of superiors), cert. denied, 450 U.S. 959, 101 S.Ct. 1418, 67 L.Ed.2d 383 (1981).
As we went on to state in Loeb v. Textron, 600 F.2d at 1012 n. 6 (citation omitted):
The reasonableness of the employer’s reasons may of course be probative of whether they are pretexts. The more idiosyncratic or questionable the employer’s reason, the easier it will be to expose it as a pretext, if indeed it is one. The jury must understand that its focus is to be on the employer’s motivation, however, and not on its business judgment.
Thus, in assessing pretext in this instance, our focus must be on the perception of the decisionmaker, i.e., whether NET perceived the plaintiff as violating the back-to-work agreement and other company policies and whether this perception was credible and reasonable, in determining whether there was a jury issue. The plaintiff has failed to present any evidence tending to show that NET did not in fact believe that he was violating the back-to-work agreement. The plaintiff did not refute any of NET’s testimony about its perception that he was violating the agreement. All of Gray’s refutation of NET’s articulated reasons related only to an explanation of why he did certain things, rather than that NET was wrong about what he did. For example, Gray claimed that he only served process as deputy sheriff for Hillsborough County during his lunch hour. He did not assert that NET was incorrect in perceiving that he served process at all hours of the day, but only that he did this service on his lunch hour regardless of what time he took it. He, thus, explains why NET exercised its decision poorly; he does not show pretext for age discrimination. Additionally, the plaintiff argues the other side of the coin: that other employees performed personal errands during working hours and were not fired. But Gray presented no direct evidence that other younger employees in a similar position performed personal errands to a similar extent and were treated differently. There is only one, general, conclusory statement that other employees did errands on company time. This sort of testimony is not enough to draw into question the credibility of NET’s perception or the reasonableness of the perception.
Similarly, Gray only explains why he spent long periods of time at the courthouse — he was doing legal research. He does not call into question NET’s perception that this was so, nor does he dispute NET’s claim that under the back-to-work agreement, Gray was required to perform his work at the office, and not elsewhere. We, therefore, conclude that the explanations Gray presents are not enough to pose a jury issue because they do not address whether NET was merely making pretextual arguments to cover its true reason for discharging Gray — age discrimination.
Looking at the evidence in the light most favorable to Gray, we cannot find that reasonable inferences could be drawn so as to allow the ADEA claim to go to the jury. While the plaintiff’s reliance on Dace argues that it is only with extreme hesitancy that cases should be withheld from the jury, we cannot agree that the jury should have been given the opportunity to consider this case. As in Dace, we have assumed that the evidence presented by the party opposing the directed verdict motion was true and have given him the benefit of all reasonable inferences. But while in Dace there was evidence from which permissible and logical inferences of age discrimination could have been drawn, here the record allows only unreasonable and speculative inferences and lends inadequate support for a charge of age discrimination. While it is important that determinations as to the weight and credibility of evidence be given to the jury, a court cannot abdicate its responsibilities under Fed.Rule Civ.Pro. 50(a) and permit a case to go to the jury when it rests only on unreasonable inferences. Accordingly, we affirm the district court’s grant of a directed verdict against the plaintiff.
Right to Counsel
The second challenge Gray brings is to the district court’s order at the June pre*257trial conference that Gray proceed to trial on July 8, 1985 even if he must appear pro se. We find no merit to this claim.
While right to counsel in the criminal and civil context are not identical, a civil litigant does have a constitutional right, deriving from due process, to retain hired counsel in a civil case. See Potashnick v. Port City Construction Co., 609 F.2d 1101, 1117-19 (5th Cir.), cert. denied, 449 U.S. 820, 101 S.Ct. 78, 66 L.Ed.2d 22 (1980) . This right has been considered so important that it has been extended to certain administrative proceedings as well. See, e.g., Goldberg v. Kelly, 397 U.S. 254, 90 S.Ct. 1011, 25 L.Ed.2d 287 (1970) (in welfare benefits termination hearing, recipient must be allowed to retain an attorney if he so desires).
The courts, considering the right to counsel in civil and administrative cases, however, have been concerned with the party’s opportunity to retain counsel and the party’s access to counsel. See, e.g., Doe v. District of Columbia, 697 F.2d 1115, 1119 (D.C.Cir.1983) (“District courts must be ... chary of issuing protective orders that restrict the ability of counsel and client to consult with one another during trial or during the preparation therefor.”); Mosley v. St. Louis Southwestern Railway, 634 F.2d 942, 945-46 (5th Cir.) (EEOC investigator erroneously denied plaintiff the opportunity to discuss a proposed settlement with retained counsel), cert. denied, 452 U.S. 906, 101 S.Ct. 3032, 69 L.Ed.2d 407 (1981) ; Potashnick, 609 F.2d at 1118 (Judge’s rule prohibiting a litigant from consulting with his attorney during breaks and recesses in the litigant’s testimony impinges upon right to retain hired counsel).
In the case at bar, we find no infringement of Gray’s due process rights. Gray was neither barred from retaining counsel nor denied access to counsel. The district court judge simply put a not unreasonable limit on the time for which he would postpone a scheduled trial while waiting for the plaintiff to retain alternate counsel. Far from denying Gray the opportunity to retain counsel, the judge gave the plaintiff a continuance of the original December trial date and allowed Gray to seek counsel from November until July. We find no abuse of discretion in the trial judge’s decision that no further delay should be tolerated and that the plaintiff be required to proceed to trial with or without counsel as of the July trial date.
Attorney’s Fees
NET claims as error on its appeal the trial court’s denial of its motion for an award of attorney’s fees from Gray in the amount of $62,258.63. NET first argues that the trial judge applied the incorrect standard in ruling attorney’s fees were not to be awarded on the ERISA claim. It then argues that, even assuming the correct standard was applied, the trial judge abused his discretion in denying the award. Similarly, NET claims that the trial judge abused his discretion in denying attorney’s fees for the wrongful discharge and ADEA claims. We find NET’s arguments without merit and affirm the decision of the court below.
Attorney’s fees are available under ERISA pursuant to 29 U.S.C. § 1132(g)(1): “In any action under this subchapter ..., the court in its discretion may allow a reasonable attorney’s fee and costs of action to either party.” In considering the application for fees, the district court applied a five factor standard which has been adopted by the Second, Third, Fifth, Eighth, Ninth, Tenth, and Eleventh Circuits. See Miles v. New York State Teamsters Conference, 698 F.2d 593 (2d Cir.), cert. denied, 464 U.S. 829,104 S.Ct. 105, 78 L.Ed.2d 108 (1983); Ursic v. Bethlehem Mines, 719 F.2d 670, 673 (3d Cir.1983); Ironworkers Local No. 272 v. Bowen, 624 F.2d 1255 (5th Cir.1980); Lawrence v. Westerhaus, 749 F.2d 494 (8th Cir.1984); Smith v. CMTA-IAM Pension Trust, 746 F.2d 587 (9th Cir.1984); Gordon v. U.S. Steel Corp., 724 F.2d 106 (10th Cir.1983); Fine v. Semet, 699 F.2d 1091 (11th Cir. 1983). This five factor standard considers (1) the degree of bad faith or culpability of the losing party; (2) the ability of such *258party to personally satisfy an award of fees; (3) whether such award would deter other persons acting under similar circumstances; (4) the amount of benefit to the action as conferred on the members of the pension plan; and (5) the relative merits of the parties’ positions.
NET asserts, however, that the standard adopted by the Seventh Circuit in Bittner v. Sadoff & Rudoy Industries, 728 F.2d 820 (7th Cir.1984) is the appropriate one to be applied in this action. The Bittner court reasoned that the five factor standard is inappropriate in cases where the defendant, rather than the plaintiff prevails, because the second, third and fourth factors have no bearing whatsoever in evaluating a case where the defendant prevails. Concluding that a prevailing defendant in an ERISA suit should be afforded the same opportunity to obtain attorney’s fees that is afforded a prevailing plaintiff, the Seventh Circuit instead adopted the Equal Access to Justice Act (hereinafter “EAJA”), 28 U.S.C. § 2412(d)(1)(A) (repealed Oct. 1, 1984), standard for determining whether prevailing defendants should collect fees:
Except as otherwise specifically provided by statute, a court shall award to a prevailing party other than the United States fees and other expenses, in addition to any costs awarded pursuant to subsection (a), incurred by that party in any civil action (other than cases sounding in tort) brought by or against the United States in any court having jurisdiction of that action, unless the court finds that the position of the United States was substantially justified or that special circumstances make an award unjust.
Thus, in an ERISA suit, the Bittner court would award a prevailing defendant attorney’s fees unless the court found that the position of the plaintiff was substantially justified or that special circumstances make an award unjust.
We decline to follow the Seventh Circuit approach and agree with the district court that the five factor standard is the appropriate test for awarding fees pursuant to ERISA, 29 U.S.C. § 1132(g)(1) with respect to both prevailing plaintiffs and prevailing defendants. The five factor test which has traditionally been employed in ERISA fee award cases is a flexible one, developed so as to give guidance to courts in interpreting the discretion to be exercised under the statute. The courts adopting this standard have emphasized that the five factors are indeed guidelines, that not every factor must be considered in each case, see, e.g., Carpenters Southern California Administrative Corp. v. Russell, 726 F.2d 1410, 1416 (9th Cir.1984), and that no one of them should be dispositive, see, e.g., Terpinas v. Seafarer’s International Union, 722 F.2d 1445 (9th Cir.1984).
The test was originally spelled out by the Tenth Circuit in Eaves v. Penn, 587 F.2d 453, 465 (10th Cir.1978), but has since been aodpted by virtually every court of appeals, as indicated above. The cases applying the standard have not been limited to those considering awards to prevailing plaintiffs, but have applied the five factor test when prevailing defendants have sought awards as well. E.g., Gordon v. U.S. Steel Corp., 724 F.2d 106 (10th Cir.1983); Sapper v. Lenco Blade, Inc., 704 F.2d 1069 (9th Cir. 1983).
We are in agreement with these other circuits that the five factor standard is the appropriate one to be applied regardless of which party prevails. Because it is not a rigid test, but rather provides general guidelines for the district court judge’s exercise of discretion, the five factor approach allows for award of fees to defendants in proper cases. We disagree with NET’s reasoning that the second, third, and fourth factors are inapplicable in prevailing defendant cases. Ability to satisfy an award, deterrence, and benefit conferred by the parties’ positions are all relevant considerations in the decision to award a fee. Even if these factors favor a prevailing plaintiff more strongly than a prevailing defendant, see Carpenters Southern, 726 F.2d at 1416; Marquardt v. North American Car Corp., 652 F.2d 715 (7th Cir. 1981), we do not believe that this result *259is out of keeping with the purposes of ERISA. While the ERISA attorney’s fees provision may not have been intended to encourage “private attorney general” type suits by providing fees almost as a matter of course as was true of the Civil Rights Attorney’s Fees provision, 42 U.S.C. § 1988, see H.R.Rep. No. 1558, 94th Cong., 2d Sess. 6-7 (1976), U.S.Code Cong. & Admin.News, 1976, p. 5908, ERISA was primarily intended to protect the interests of plan beneficiaries and participants; therefore, such a “bias” in the standard does not thwart the legislative injunctive that attorney’s fees may be awarded to “either” party, in the court’s discretion. Factors may simply not weigh equally where defendants rather than plaintiffs seek fees.
Furthermore, we have reservations about the transferability of the EAJA attorney’s fees standard to the ERISA context. While perhaps it is true that both the ERISA and EAJA fee provisions were intended to trod a middle ground in the award of attorney’s fees, Bittner, 728 F.2d at 830 (ERISA took middle ground approach); United States v. Yoffe, 775 F.2d 447 (1st Cir.1985) (EAJA took middle ground approach), one middle ground is not necessarily the same as another. The EAJA provision is, in fact, a one-way statute allowing recovery only against the United States. The considerations motivating the development of that standard cannot, thus, be assumed to apply equally in ERISA cases. EAJA attorney’s fees cases focus almost exclusively on the “reasonableness” of the government’s position in the litigation. See Spencer v. N.L.R.B., 712 F.2d 539, 557-58 (D.C.Cir.1983), cert. denied, 466 U.S. 936, 104 S.Ct. 1908, 80 L.Ed.2d 457 (1984). Concerns about ability to pay, deterrence, and benefit simply are not as relevant when recovery is sought against the government. The Congress decided the government should pay in these sorts of cases; because the government assumedly has competent counsel, whether a suit is harassing can more easily be adjudged from its litigation position; and, again, Congress simply chose to assume that non-United States parties provide a benefit by bringing or defending EAJA suits while the United States only does so where it is substantially justified in its litigation position.
The five factor test does incorporate a “reasonableness” component, but simply makes other considerations relevant as well. A multi-faceted approach is more desirable when fees are to be assessed against parties other than the United States. The five factor standard, rather than a strict “substantially justified” approach, will best guide a judge’s discretion in keeping with the policies of ERISA and the fact that individual plan participants may be the plaintiffs. Accordingly, we affirm the district court’s use of the five factor test.
Given that the correct standard was applied, our review of the trial judge’s denial of attorney’s fees is limited to a review for abuse of discretion. E.g., Evans v. Bexley, 750 F.2d 1498 (11th Cir.1985). No such abuse can be found here.
We note first that in this circuit, an award of attorney’s fees under ERISA is not virtually automatic, but remains discretionary with the trial judge. See Lodge v. Shell Oil Co., 747 F.2d 16, 20 (1st Cir. 1984). What NET argues is simply that because Gray’s position was without merit, as reflected by the trial judge’s grant of a directed verdict, attorney’s fees are appropriate here. NET does try to infer that Gray acted in bad faith, but in fact, NET simply reasserts its claim that Gray’s suit was totally meritless, without pointing to any other circumstances indicating that Gray had knowledge that his suit was without, basis, but proceeded anyway. As we will not equate failure of proof with bad faith, we are right back to considering the reasonableness of Gray’s ERISA claim. As indicated above, no one of the five factors should necessarily be dispositive in awarding a fee. Unreasonableness in a party’s litigation position does not require a fee award. Further, there is some evidence in the record of NET’s withholding of pension benefits from the plaintiff, so *260that NET’S accusation that Gray’s claim was totally lacking in legal and factual merit is somewhat exaggerated. We conclude that the trial judge had the opportunity to observe the proceedings, parties, and claims throughout the litigation, considered the five factors, and did not abuse his discretion in denying fees on the ERISA claim.
Similarly we find no abuse of discretion in the trial court’s denial of fees to NET on the wrongful discharge and ADEA claims. Applying relevant New Hampshire law, see McGinty v. Beranger Volkswagen, Inc., 633 F.2d 226 (1st Cir.1980), attorney's fees on the wrongful discharge claim may only be awarded if the opposing party acted in bad faith in the litigation. Harkeem v. Adams, 117 N.H. 687, 691-92, 377 A.2d 617, 619-20 (1977). As with the ERISA claim, we find no abuse of discretion in the trial court’s finding of no bad faith; NET has not shown us that Gray persisted in advocating a “patently” merit-less position. And finally, under the bad faith requirements for award of fees under Alyeska Pipeline Service Co. v. Wilderness Society, 421 U.S. 240, 258-59, 95 S.Ct. 1612, 1622, 44 L.Ed.2d 141 (1975), applicable to the ADEA claim,1 NET has again failed to show that abuse of discretion occurred here.
Affirmed.
BAILEY ALDRICH, Senior Circuit Judge, dubitante.
Perhaps I read more into the court’s opinion than I should, and, if so, I should keep silent. However, just in case, my appraisal of what it is saying is that although the plaintiff has made a prima facie case requiring defendant to show a proper reason for his discharge, defendant has met it as matter of law even though plaintiff’s evidence would warrant a finding that the given reason was, in fact, untrue.
If a defendant has to give a reason, and offers one, can it be ruled as matter of law that the jury has to credit it? It may be that somehow this is so here because of plaintiff’s ultimate burden of showing pretext, but if, as it would seem, the court is saying that plaintiff cannot make a case for the jury by affirmatively showing the reason was untrue, I think defendant has escaped too easily. This would seem reminiscent of the days when some courts held that a presumption disappeared when evidence was introduced against it even though the jury did not believe the evidence.
. Under ADEA, an award of attorney’s fees is only authorized for plaintiffs. 29 U.S.C. § 262(b). However, defendants may obtain fees under Alyeska.