The issue before us in this case is whether the trial judge abused his discretion when he denied the nonparty insurance company’s postjudgment motion to intervene as a defendant for the purpose of pursuing an appeal of the plaintiffs’ jury verdict.
*57Briefly, a jury found the defendant, O’Connor Café of Worcester, Inc., doing business as Leitrim’s Pub (Leitrim’s), liable in this dramshop liability suit and awarded the plaintiffs Lynn and Geraldine Bolden (Boldens) over twenty million dollars in damages. Pursuant to a postjudgment settlement agreement between the parties, Leitrim’s agreed to assign to the Boldens any rights or claims it might have against its liability insurer, the Liquor Liability Joint Underwriting Association of Massachusetts (LLJUA). In exchange, the Boldens agreed not to execute the judgment against Leitrim’s. Thereafter, the Bold-ens served the LLJUA with a G. L. c. 93A demand letter which alleged that pretrial settlement attempts by the LLJUA, which provided a defense to Leitrim’s, were made in bad faith. Faced with the prospect of having to defend itself in a G. L. c. 93A lawsuit that would arguably use the jury verdict in this dram-shop case as a fulcrum, the LLJUA moved to intervene as a party in order to pursue the appeal of the verdict against Leitrim’s. The LLJUA appeals from the denial of its motion.
Facts and procedural history. We recite additional facts to give context to the issue presented. On March 10, 1991, Lynn Bolden suffered severe brain damage after being injured in a motor vehicle accident. Lynn Bolden was a passenger in a car driven by Maria Little. Lynn Bolden and her mother Geraldine Bolden (who is also her legal guardian) filed a personal injury suit against Leitrim’s in 1994. The Boldens alleged that, prior to the accident, employees of Leitrim’s had negligently served alcohol to Little after they knew or should have known that Little was intoxicated.
When the accident occurred, Leitrim’s had coverage under a liquor liability insurance policy issued by the LLJUA with a policy limit of $100,000.3 As Leitrim’s insurer, the LLJUA agreed to and did defend Leitrim’s in the Boldens’ suit. Before the case went to the jury in August, 1997, the LLJUA offered *58the Boldens $5,000 to settle their claims. The Boldens rejected this offer. After the jury began its deliberations, the LLJUA increased its offer to $15,000. That figure was also rejected.4 On August 21, 1997, the jury returned a verdict for the Boldens, awarding them more than twenty million dollars.5
Eight days after the verdict, the LLJUA issued a check to the Boldens for $100,000, the limits of Leitrim’s liability policy. The LLJUA made this payment despite the fact that the Boldens had declined the LLJUA’s proposal to accept the check in full settlement of their claim against Leitrim’s.
In the weeks following the verdict, Leitrim’s counsel became increasingly concerned as to whether the LLJUA would fund Leitrim’s appeal of the verdict.6 There is no indication in the record as to when the LLJUA actually informed Leitrim’s that it would fund the appeal, but notice of appeal was in fact filed on October 16, 1997, after the trial judge denied Leitrim’s motion for judgment notwithstanding the verdict.
Shortly before the notice of appeal was filed, Leitrim’s and the Boldens executed an assignment and release agreement on October 8, 1997, whereby Leitrim’s assigned to the Boldens any rights or claims it might have against the LLJUA and its counsel in exchange both for the Boldens’ promise not to execute on the judgment and for their release of Leitrim’s from *59any additional claims against it.7 Leitrim’s also agreed to cooperate with the Boldens in any suit brought to enforce any right assigned them by Leitrim’s.
On December 1, 1997, the Boldens, now standing in Leitrim’s shoes, served the LLJUA with a G. L. c. 93A demand letter. The Boldens alleged that the LLJUA violated G. L. c. 176D, § 3(9)0, and G. L. c. 93A, § 2(a), in that the LLJUA failed to offer its policy limits as a settlement after Leitrim’s liability had become reasonably clear,8 and further alleged that the LLJUA’s settlement attempts were made in bad faith.9 The Boldens demanded over twenty-one million dollars. The *60LLJUA claims it was unaware that Leitrim’s had assigned its rights until it received the demand letter.
The Boldens moved to dismiss Leitrim’s appeal on March 27, 1998. They submitted an affidavit attesting that Leitrim’s did not oppose this motion. The LLJUA then filed an emergency motion to intervene for the purpose of prosecuting Leitrim’s appeal, which the Boldens opposed. The LLJUA sought to intervene either as of right or permissively, pursuant to Mass.R. Civ.P. 24(a)(2) and (b)(2), 365 Mass. 769-770 (1974), respectively. In arguing that it had a right to intervene, the LL-JUA asserted that it had an actual and significant interest in the outcome of the dramshop liability tort suit appeal, since it now faced liability in a separate G. L. c. 93A suit in which it would be unable to reopen the issue of Leitrim’s liability. In terms of permissive intervention, the LLJUA argued that it had a claim or defense in common with the dramshop liability suit, since the issue of Leitrim’s liability was present in the G. L. c. 93A suit as well.
The trial judge heard argument on both the LLJUA’s motions to intervene and on the Boldens’ motion to dismiss the appeal on April 3, 1998. Later that month, the Boldens filed a G. L. c. 93A suit alleging the same claims previously asserted in their demand letter. In August, 1998, the trial judge denied the LL-JUA’s motion and granted the Boldens’ motion to dismiss the appeal. This appeal followed.10
Intervention as of right. The LLJUA moved to intervene as of right under Mass.R.Civ.P. 24(a), which reads in pertinent part: *61See Massachusetts Fedn. of Teachers v. School Comm. of Chelsea, 409 Mass. 203, 205 (1991).
*60“Upon timely application anyone shall be permitted to intervene in an action ... (2) when the applicant claims an interest relating to the property or transaction which is the subject of the action and he is so situated that the disposition of the action may as a practical matter impair or impede his ability to protect that interest, unless the applicant’s interest is adequately represented by existing parties.”
*61Otherwise put, the proposed intervener as of right must satisfy four criteria: (1) the application must be timely; (2) the applicant must claim an interest relating to the property or transaction which is the subject of the litigation in which the applicant wishes to intervene; (3) the applicant must show that, unless able to intervene, the disposition of the action may, as a practical matter, impair or impede his ability to protect the interest he has; and (4) the applicant must demonstrate that his interest in the litigation is not adequately represented by existing parties. The LLJUA contends that it meets each of these criteria. We review rule 24(a)(2)' decisions under an abuse of discretion standard. See Cosby v. Department of Social Servs., 32 Mass. App. Ct. 392, 395 (1992).
As to the first, timeliness, we observe that postjudgment motions to intervene, whether as of right or permissive, are seldom timely. See Motor Club of America Ins. Co. v. McCroskey, 9 Mass. App. Ct. 185, 187-188 (1980); Cruz Mgmt. Co. v. Thomas, 417 Mass. 782, 785 (1994). The proposed postjudgment intervener accordingly must not only justify its failure to intervene at an earlier stage of the action, but must also establish that it has not just an interest, but a compelling one, in the litigation. See Cruz Mgmt., 417 Mass. at 786.
In the circumstances here, we think the LLJUA can justify its failure to intervene earlier in the dramshop liability suit, when it was providing Leitrim’s, its insured, with a defense. Put simply, the LLJUA’s asserted interest in that tort litigation did not arise until shortly before it sought to intervene. Whether that asserted interest is a compelling interest, however, requires a good look at the second and third criteria for intervention as of right, i.e., whether the LLJUA has a cognizable, let alone compelling, interest in the dramshop litigation and whether it can adequately protect that interest absent intervention. The question of timeliness in this sense merges with the question of what interest the LLJUA has in the property or transaction that is the subject of the dramshop liability litigation.
Much hand-wringing has been done since the enactment of rule 24(a)(2) by judges who are called upon to decide whether its requirements have been satisfied in a particular case. Nowhere is the difficulty greater than in determining whether a proposed intervener has an interest sufficient to entitle it to *62intervene. The rule itself does not articulate explicit criteria for determining the sufficiency of the asserted interest. Accordingly, we have agreed that a “flexible, rather than rigid approach is indicated” and that “the requirement should be viewed as a prerequisite rather than relied upon as a determinative criterion for intervention.” Cosby v. Department of Social Servs., 32 Mass. App. Ct. at 395-396, quoting from Smuck-v. Hobson, 408 F.2d 175, 179 (D.C. Cir. 1969) (en banc).
The Supreme Court has stated that the interest in the litigation must be “significantly protectable,” Donaldson v. United States, 400 U.S. 517, 531 (1971), and must be sufficiently direct and immediate to justify the intervention, see Cascade Natural Gas Corp. v. El Paso Natural Gas Co., 386 U.S. 129, 154 (1967). The interest does not suffice if it is remote or contingent. See Restor-A-Dent Dental Labs. v. Certified Alloy Prods., 725 F.2d 871, 874 (2d Cir. 1984); Travelers Indem. Co. v. Dingwell, 884 F.2d 629, 638 (1st Cir. 1989).11 But see Securities & Exchange Commn. v. Flight Trans. Corp., 699 F.2d 943, 948 (8th Cir. 1983). Nor may the interest be a tangential or collateral one. See Purcell v. BankAtlantic Fin. Corp., 85 F.3d 1508, 1513 (11th Cir. 1996); United States v. Union Elec. Co., 64 F.3d 1152, 1161 (8th Cir. 1995).
In the end, however, there is “no convenient rule of thumb” which we can employ, Mayflower Dev. Corp. v. Dennis, 11 Mass. App. Ct. 630, 635 (1981), and we are thrown back upon the need for a practical, case-specific, fact-intensive analysis. As a result, other appellate decisions are generally only of limited assistance. In this regard, we agree entirely with the First Circuit’s view that “[t]he doctrinal ‘rules’ are so general as to provide very little help, and the diversity of facts and interests that may be affected by litigation today is so broad as to make any simple formula difficult, if not impossible, to contrive” (citation omitted). Cotter v. Massachusetts Assn. of Minority Law Enforcement Officers, 219 F.3d 31, 34 (1st Cir. 2000).
The LLJUA describes its asserted interest in the property or transaction which is the subject of the dramshop liability litiga*63tian along these lines. It should be allowed to pursue the appeal, it suggests, because it should have the opportunity to overturn the jury verdict which found Leitrim’s liable for having negligently served alcohol to a patron when Leitrim’s employees knew or should have known that patron was intoxicated. Were the LLJUA able to appeal successfully, the argument goes, with the result that Leitrim’s was no longer hable, then the LLJUA would be home free in the G. L. c. 93A lawsuit it faces for unfair settlement practices. The LLJUA explains it thus: it could not be faulted in the G. L. c. 93A case for not having made fair offers in the tort case once Leitrim’s liability became reasonably clear because Leitrim’s never was hable at all and its liability, accordingly, could never have been reasonably clear.12 See Lazaris v. Metropolitan Property & Cas. Ins. Co., 428 Mass. 502, 505 (1998).
The trial judge saw the LLJUA’s asserted interest somewhat differently. He concluded that the LLJUA’s interest in the appeal was to prevent the Boldens, standing in the insured Leitrim’s shoes, from preserving the large damage award they received in the dramshop habihty suit and later asserting it as the amount that might be used to measure damages in the G. L. c. 93A lawsuit against the insurer for unfair settlement practices.
The judge determined that this interest was merely contingent, insofar as the Boldens not only had to file a c. 93A action13 and prevail on it by establishing all necessary elements to prove that the LLJUA had failed to effectuate prompt, fair, and equitable settlement of the Boldens’ claim when Leitrim’s liability had become reasonably clear, but they also had to prevail on their allegation that the LLJUA had acted in bad faith, and that such bad faith settlement practice resulted in the foreseeable consequence of the excess jury verdict. Only if that were so and the jury verdict deemed to be the measure of the Boldens’ actual damages would the LLJUA face the music. As a merely *64contingent interest, the judge thought it insufficient to satisfy the requirement of rule 24(a)(2).
In considering whether the interests asserted by the LLJUA pass muster under the rule, it bears mention that the LLJUA no longer faces any direct liability of its own as Leitrim’s insurer in the present dramshop liability case since it has long since paid the policy limits to the Boldens. Nor does Leitrim’s, its insured, face fiirther liability, having reached a settlement with the Boldens in regard to that part of the judgment which was in excess of the policy limits. The asserted LLJUA interests are, in essence, concerns as to its exposure to potentially massive liability, not in this tort suit, but in the separate G. L. c. 93A action it faces for unfair settlement practices arising from its own actions in the dramshop liability tort suit. It is that potential liability which the LLJUA wishes to limit or eliminate by appeal in this tort case. It is important to note in this regard, however, that no one suggests that the dramshop suit would have res judicata or collateral estoppel effect on the G. L. c. 93A case.
The LLJUA urges us to view its interest in the dramshop liability appeal as a direct one, particularly insofar as its success on appeal for Leitrim’s might preclude any liability for itself in the G. L. c. 93A case. However, since it appears to be only the limited and potentially helpful collateral effect of the dramshop liability tort action upon the G. L. c. 93A case — and not any further liability on the part of the insurer LLJUA or the insured Leitrim’s in this tort action — which causes the LLJUA to be interested in the dramshop liability tort action, the LLJUA’s interest is more fairly described as collateral than as direct. The LLJUA cites no comparable case, and we are aware of none, in which an entity was permitted to intervene as of right to defend its interest in another lawsuit.
We think the closest case that supports our characterization of the LLJUA’s asserted interests as collateral is Purcell v. Bank-Atlantic Fin. Corp., 85 F.3d 1508 (11th Cir. 1996). In Purcell, the plaintiffs alleged securities violations by the defendants in a securities action which ultimately resulted in a large jury verdict. The defendants then sued a television broadcaster for having libelously described the defendants’ securities activities. The broadcaster wished to use the collateral effect of the jury verdict in the securities suit as a complete defense in the libel suit, but the parties to the securities action settled and that judgment was vacated. The broadcaster was not entitled to intervene in the *65securities suit to oppose the vacatur notwithstanding its undeniable but collateral interest in the securities suit jury verdict. As here, the broadcaster had wished to intervene in the first suit solely to defend its interest in the second suit. 85 F.3d at 1512-1513.
While describing the LLJUA’s asserted interests as collateral rather than direct suggests their insufficiency under rule 24(a)(2), the very magnitude of the LLJUA’s potential exposure14 in the G. L. c. 93A suit still must give us pause and cause us to eschew labels in favor of further analysis. The LLJUA’s asserted interests15 in appealing — to overturn its insured’s liability and to *66reduce or eradicate the excess judgment — appear at first glance to loom quite large, particularly if one dwells on the possibility that the twenty million dollar jury verdict against the insured Leitrim’s in the dramshop case could be asserted as the measure of damages for doubling or trebling in the G. L. c. 93A case against the insurer LLJUA. But all may not be as it first seems, and one way to test the nature and depth of the asserted interests is to examine what might happen to them if the LLJUA were unable to intervene and pursue the appeal. This takes us to the third criterion — whether the LLJUA has shown that, as a practical matter, absent intervention, its ability to protect the asserted interests might be impaired or impeded.
The LLJUA says that its appellate goal is to eradicate Leitrim’s dramshop liability. It suggests that if it is unable to attack on appeal the asserted evidentiary errors and insufficiency of the evidence as to whether Leitrim’s knew or had reason to know of the patron driver’s intoxication when it served her alcohol, it will be forced at the G. L. c. 93A trial to re-litigate the dram-shop liability case. This is not quite so. We turn to an analysis of the issues to be decided in the G. L. c. 93A unfair settlement practices suit in order to determine whether the LLJUA might be impaired in its ability to protect itself there because it was unable to appeal here.
The issues to be determined in the G. L. c. 93A claim do not in fact concern Leitrim’s actual liability. The LLJUA’s settlement practices could be entirely without fault and the LLJUA accordingly immune from G. L. c. 93A liability, yet the dram-shop jury could still have found Leitrim’s liable for dramshop liability. The LLJUA need not show at the G. L. c. 93A trial that Leitrim’s was without fault. What the LLJUA must instead do there is show that its settlement offers in the dramshop case were reasonable and made in good faith, given its own knowledge at the time of the relevant facts and law concerning the Boldens’ claim. See Parker v. D’Avolio, 40 Mass. App. Ct. 394, 395 (1996). The resolution of the G. L. c. 93A claim, *67including the issue of bad faith, will depend upon a factual determination of the LLJUA’s knowledge and intent. See Williams v. Gulf Ins. Co., 39 Mass. App. Ct. 432, 436 (1995).16 Accordingly, the LLJUA will not need to attack collaterally the dramshop liability judgment in order to prove in the G. L. c. 93A case that Leitrim’s was not actually liable and therefore that the LLJUA is also not liable for unfair settlement practices. Rather, the LLJUA need only demonstrate that Leitrim’s liability was not “reasonably clear” to the LLJUA, not to the jury that heard the dramshop liability case.
Otherwise put, notwithstanding what the dramshop liability jury concluded about Leitrim’s liability, what matters in the G. L. c. 93A case is whether the LLJUA reasonably believed that Leitrim’s liability was not clear, or was unreasonable in holding that belief. “So long as the insurer acts in good faith, the insurer is not held to standards of omniscience or perfection; it has leeway to use, and should consistently employ, its honest business judgment.” Peckham v. Continental Cas. Ins. Co., 895 F.2d 830, 835 (1st Cir. 1990). Because Leitrim’s actual liability will not be “retried” in the G. L. c. 93A suit, we conclude that the LLJUA will be able to defend itself without impairment on the relevant issues in the G. L. c. 93A suit.17
It bears mention as well that the issues that the LLJUA is burning to raise on appeal in the dramshop case to protect its interests in the G. L. c. 93A case do not signal the kind of grave injustice occurring at a jury trial that cries out to be rectified by an appellate court. The LLJUA’s argument is notably vague and skimpy on the details of the appellate issues it wishes to pursue, suggesting only that they involve evidentiary errors and insufficiency of the evidence as to a key element of the tort. It is relatively rare for evidentiary errors to result in a reversal in a civil action. As to the sufficiency of the evidence, an experienced trial judge has already had three opportunities to consider the adequacy of the evidence — on the motion for directed verdict *68at the close of plaintiffs’ evidence, the renewal of that motion at the end of all the evidence, and on the posttrial motion for judgment notwithstanding the verdict. We do not imply that either category of appeal is hopeless, only that the LLJUA has described no specific error or insufficiency that supports serious prospects for reversal on appeal. Apart from this, we are mindful that the very ability of the LLJUA to intervene and appeal might by itself help the LLJUA’s cause by providing it with the bit of leverage so often useful to favorable settlement. That the delay and uncertainty incumbent on appeal may bring such advantage in their wake is not the type of interest courts may consider.
Although the LLJUA stresses the liability aspect as its chief interest in the G. L. c. 93A case and does not dwell on the desirability of its being able to reduce or eradicate the large judgment, this is nonetheless a point we should consider. Were the Boldens successful in establishing in the c. 93A action that the LLJUA’s bad faith settlement practices foreseeably caused the excess verdict to enter in the dramshop case, that verdict arguably would be the measure of damages to be doubled or trebled.
It appears at least somewhat unsettled, however, whether such an underlying judgment remains the basis for “actual damages” under G. L. c. 93A, § 9(3), if, as here, a settlement relieving the insured from liability from the excess judgment is reached after judgment. In DiMarzo v. American Mut. Ins. Co., 389 Mass. 85, 94-95 (1983), a G. L. c. 93A, § 9, case decided before that section was amended in 1989, the court reiterated that “the entry of a judgment against a person constituted a loss of money” as required by G. L. c. 93A, § 9(1), and that “[t]he loss does not turn on whether the judgment has been satisfied.” The 1989 amendment to § 9(3), however, has been interpreted as imposing certain limitations on the concept. See Cohen v. Liberty Mut. Ins. Co., 41 Mass. App. Ct. 748, 755 (1996), and note 9, supra. See also Clegg v. Butler, 424 Mass. 413, 425-426 (1997) (“judgment” “does not encompass damages awarded as part of a settlement” made before trial of underlying claim; multiple damages limited to interest lost on money withheld wrongfully by insurer); Bonfiglio v. Commercial Union Ins. Co., 411 Mass. 31, 37 (1991), S.C., 412 Mass. 612 (1992) (arbitrator’s award not “judgment”). Compare Miller v. Risk Mgmt. Foundation of Harvard Med. Insts., Inc., 36 Mass. App. Ct. *69411, 420 n.15 (1994); Metropolitan Property & Cas. Ins. Co. v. Choukas, 47 Mass. App. Ct. 196, 202 (1999). If the judgment in the present case were not the basis for a G. L. c. 93A damages award due to the agreement by the Boldens not to execute on the judgment, the LLJUA arguably would be responsible only for the ascertainable losses suffered by Leitrim’s which were the foreseeable consequence of the LLJUA’s unfair conduct. See Cohen, 41 Mass. App. Ct. at 755. All of this, however is not for us to settle today, nor would the point for that matter be resolved in the appeal of the dramshop case. It is rather one for the LLJUA to raise if it wishes at trial in the G. L. c. 93A action and, if necessary, on an appeal from that judgment.18
To sum up, then. We recognize that the LLJUA has interests that it understandably wishes to protect, but it has not demonstrated that they are sufficiently direct or compelling so as to entitle it to intervene in the dramshop case as a matter of right. We are further persuaded that, while the LLJUA may not be able to eradicate the G. L. c. 93A action ab initia by virtue of appellate victory in the dramshop case, the LLJUA’s ability to protect its asserted interests in avoiding liability in the G. L. c. 93A case will not likely be impeded or impaired absent intervention. We think the LLJUA is able to protect those interests adequately, “in the old-fashioned way,” Purcell, 85 F.3d at 1514, in the course of the G. L. c. 93A trial.19 In so saying, we also recognize that this is a complex set of circumstances and that reasonable minds could differ in assessing whether the LLJUA adequately satisfies the criteria for intervention as of right. The standard of review by which we are bound is whether the trial judge abused his discretion in denying the LLJUA’s rule 24(a)(2) motion. Cosby v. Dept. of Social Servs., 32 Mass. App. Ct. at 395.20 It is evident from his memorandum that the *70experienced trial judge gave careful consideration to all the correct factors before concluding that the LLJUA had not shown a compelling interest entitling it to intervene. We are not prepared to substitute our judgment for his or to say that he abused his discretion.
Permissive intervention. The LLJUA also appeals the judge’s denial of the LLJUA’s motion to intervene by permission. A judge may permit intervention “when an applicant’s claim or defense and the main action have a question of law or fact in common.” Massachusetts Fedn. of Teachers, 409 Mass. at 209, quoting from Mass.R.Civ.P. 24(b)(2), 365 Mass. 769 (1974). A' judge must also consider “whether allowing intervention might delay or prejudice the adjudication of the rights of the [original] parties.” Ibid. Mass.R.Civ.P. 24(b). Permissive intervention is wholly discretionary with the trial judge and the judge’s decision will be reversed only for a clear abuse of discretion. Massachusetts Fedn. of Teachers, 409 Mass. at 209. Under rule ' 24(b)(2), a judge may permit intervention, but is not obliged to do so. See Attorney Gen. v. Brockton Agric. Soc., 390 Mass. 431, 435 (1983).
The LLJUA argues here, as it did before the trial judge, that the issue of Leitrim’s liability included questions of law or fact common to both the present case and the G. L. c. 93A action. The judge disagreed, observing that the LLJUA’s settlement efforts under the circumstances were at issue in the G. L. c. 93A case, not Leitrim’s liability. The judge did not abuse his considerable discretion.
As discussed at length above, we concur in thinking that the question of Leitrim’s actual liability is not at the heart of the G. L. c. 93A suit, and that the LLJUA has presented no common question of law or fact within the present case, as required by rule 24(b)(2). Compare with Fabiano v. Boston Redev. Authy., 49 Mass. App. Ct. 66, 76 (2000) (all of proposed intervener’s defenses went to the heart of original claims brought by plaintiff, *71who was party opposing motion to intervene).21 Furthermore, we fail to see how the issue of Leitrim’s liability in the G. L. c. 93A action is a “claim or defense” on the part of the LLJUA in the present action for the purposes of rule 24(b)(2), since neither Leitrim’s nor the LLJUA faces any further liability in the present case requiring a defense and no counterclaims have been brought against the Boldens.22
Moreover, the trial judge also determined that if the Boldens and Leitrim’s were forced to engage unwillingly in the appellate process after settlement in order that the LLJUA could avoid liability, the rights of the original parties would be unduly delayed and prejudiced. We agree. In Purcell, 85 F.3d at 1514, the court viewed the proposed intervener’s assertion that its intervention would not prejudice the original parties as “woefully unconvincing,” since the intervener’s interest in blocking the vacatur of the judgment was at odds with the wishes of the parties. The LLJUA’s argument that its intervention would not be prejudicial to the original parties is equally unconvincing. The Boldens and Leitrim’s have entered into a settlement agreement that is mutually beneficial to both parties and that includes a provision precluding Leitrim’s from appealing the judgment. To force them to await the result of the appeal serves only the LLJUA’s interest, not theirs. Compare with Cruz Mgmt. Co. v. Thomas, 417 Mass. at 786 (permissive intervention would not prejudice party opposing motion); Fabiano, 49 Mass. App. Ct. *72at 76 (permissive intervention actually would be of benefit to party opposing motion). There was no abuse of discretion.
Order denying motion to intervene affirmed.
The LLJUA is a nonprofit, “statutorily mandated agglomeration of every personal liability insurer operating in the Commonwealth.” See Jimmy’s Diner, Inc. v. Liquor Liab. Joint Underwriting Assn. of Mass., 410 Mass. 61, 64-65 (1991). The LLJUA was created by the Legislature in 1985. See St. 1985, c. 223, §§ 1-18. It provides “liquor legal liability insurance to sellers and distributors of alcohol who were previously unable to obtain liability insurance in the private market.” See Liquor Liab. Joint Underwriting Assn. of Mass. v. Hermitage Ins. Co., 419 Mass. 316, 318 (1995). “Liquor legal liability insurance” is defined as “insurance coverage against the legal liability of the insured and against loss, damage, or expense incident to a claim arising *58out of death or injury to any person as a result of negligence in the distributian, sale or serving of alcohol by any licensee.” St. 1985, c. 223, § 1.
The Boldens claim that they demanded the $100,000 policy limit and would have accepted this figure had it been offered. In addition, they claim that Leitrim’s requested that the LLJUA settle the suit before it went to the jury. The LLJUA disputes both assertions.
This figure includes prejudgment interest.
A letter dated August 29, 1997, sent by an LLJUA senior claims representalive to Leitrim’s counsel, indicated that the LLJUA would fund the appeal if the appeal appeared to be meritorious. Leitrim’s counsel responded with a letter which stated that he thought that the LLJUA was “chomping at the bit” to pursue the appeal. In another letter, dated September 12, 1997, the LLJUA stated that Leitrim’s counsel had misunderstood their conversations, and in fact, only one individual involved in the decision to appeal was “chomping at the bit.” This letter further stated that all parties involved would make a decision whether an appeal would be made. Finally, a letter sent by the LLJUA to Leitrim’s counsel, dated October 16, 1997, indicated that the LLJUA would file the notice of appeal and pay for the trial transcript. This letter also stated that after reviewing the transcript and discussing the issue with corporate counsel, the LLJUA would make a determination as to whether to fund Leitrim’s appeal. (The letters discussed in this footnote were included in a supplemental record appendix filed by the Boldens with their brief.)
An agreement to enter judgment for the plaintiff in excess of policy limits coupled with a covenant not to execute in favor of the insured defendant does not invalidate an accompanying assignment of the defendant’s right to sue her insurers for negligence. See Campione v. Wilson, 422 Mass. 185, 190-194 (1996). In Campione, the court recognized that such agreements risk collusion between the injured party and the tortfeasor, but observed that this risk is reduced by the plaintiff’s burden of proof in negligence claims. 422 Mass. at 193. “Further, it is open to the defendants to suggest the issue of collusion, and if a satisfactory basis exists, to argue it to the fact finder.” Id. at 194.
General Laws c. 176D, § 3(9)(f), states that an insurance company’s failure “to effectuate prompt, fair and equitable settlements of claims in which liability has become reasonably clear” is an unfair practice. See Lazaris v. Metropolitan Property & Cas. Ins. Co., 428 Mass. 502, 505 (1998). General Laws c. 93A, § 2(a), states that “unfair or deceptive acts or practices in the conduct of any trade or commerce are hereby declared unlawful.” Those claiming injury by virtue of an insurance practice prohibited by G. L. c. 176D, § 3(9)(/), may sue under G. L. c. 93A. See Metropolitan Property & Cas. Ins. Co. v. Choukas, 47 Mass. App. Ct. 196, 197 n.1 (1999), citing Van Dyke v. St. Paul Fire & Marine Ins. Co., 388 Mass. 671, 675 (1983).
“Good faith” has been defined as the insurer making settlement decisions without regard to the policy limits and the insurer’s “exercise of common prudence to discover the facts as to liability and damages upon which an intelligent decision may be based.” Hartford Cas. Ins. Co. v. New Hampshire Ins. Co., 417 Mass. 115, 119 (1994), quoting from Murach v. Massachusetts Bonding & Ins. Co., 339 Mass. 184, 187 (1959). Further, G. L. c. 93A, § 9(3), states, in relevant part, that “if the court finds for the petitioner, recovery shall be in the amount of actual damages or twenty-five dollars, whichever is greater; or up to three but not less than two times such amount if the court finds that the use or employment of the act or practice was . . . made in bad faith with knowledge or reason to know that the act or practice complained of violated said section two. . . . For the purposes of this chapter, the amount of actual damages to be multiplied by the court shall be the amount of the judgment on all claims arising out of the same and underlying transaction or occurrence.” In Cohen v. Liberty Mut. Ins. Co., 41 Mass. App. Ct. 748, 755 (1996), we stated that even when bad faith is found in a G. L. c. 176D/93A suit, the “actual” damages are those losses which were the foreseeable *60consequence of the insurer’s unfair or deceptive conduct. Thus, it appears that, in the present situation, only if the fact finder in the Boldens’ G. L. c. 93A lawsuit were to conclude that the twenty million dollar judgment in the dramshop liability tort suit was both the measure of Leitrim’s actual damages and the foreseeable consequence of a bad faith unfair settlement practice by the LLJUA, would the underlying judgment be doubled or trebled as damages.
“Leitrim’s did not file an appellate brief.
The Reporters’ Notes to Mass.R.Civ.R 24 state that rule 24(a) is identical to Federal rule 24(a). Reporters’ Notes to Mass.R.Civ.P 24(a), Mass. Ann. Laws, Rules of Civil Procedure, at 232 (West 1992). Therefore, “the adjudged construction . . . given to the Federal [rule] is to be given to our [rule], absent compelling reasons to the contrary or significant differences in content.” Rollins Envtl. Servs., Inc. v. Superior Ct., 368 Mass. 174, 179-180 (1975).
The test used to determine whether a defendant’s liability became “reasonably clear” is an objective one; it “calls upon the fact finder to determine whether a reasonable person, with knowledge of the relevant facts and law, would probably have concluded, for good reason, that the insurer was liable to the plaintiff.” Demeo v. State Farm Mut. Auto. Ins. Co., 38 Mass. App. Ct. 955, 956-957 (1995).
The LLJUA argues that the judge’s memorandum of decision indicates that he was unaware that the Boldens’ c. 93A suit had been filed. However, the judge knew of the Boldens’ demand letter and further stated that, even if a suit was filed imminently, this would not alter his opinion.
The Boldens argue that because the LLJUA’s interest in avoiding liability is solely economic, it is insufficient to satisfy the requirements of rule 24(a)(2). However, we have held previously that an economic interest is sufficient. See Mayflower Dev. Corp. v. Dennis, 11 Mass. App. Ct. at 636 (equitable lien on plaintiff’s potential recovery satisfied interest requirement).
The LLJUA also maintains that it has a direct interest that arises from the liability insurance policy itself. The policy states that the LLJUA “shall have the right and duty to defend any suit against the Insured seeking . . . ‘damages,’ even if the allegations of the suit are groundless, false, or fraudulent.” See Jimmy’s Diner, Inc., 410 Mass. at 62. The LLJUA, however, offers little in the way of appellate argument on this point; indeed, only one page of the actual policy is included in the record appendix, and none of the provisions on that page have any bearing on the issues at hand. Compare Aetna Cas. & Sur. Co. v. Sullivan, 33 Mass. App. Ct. 154, 155-156 (1992) (relevant portion of the insurance policy was before the court). We can thus accord this contention only the limited consideration it merits.
“[T]he source of the duty to defend is the contractual agreement.” Id. at 155. However, the LLJUA points us to neither a contractual provision nor any authority that translates the duty to defend into a right to insist on an appeal after the policy limits have been paid to a claimant and the excess judgment has been settled by the parties. Nor does the LLJUA point us to any policy provision that might give it the right to prevent its insured from settling an actual or possible excess judgment against it. Compare id. at 155-156 (policy stated that insurer had “right to settle any claim or lawsuit as we see fit”). At any rate, we observe that “[t]he restriction on the insured’s right to settle does not apply to a settlement by the insured of his potential excess-liability, since the insured has a right to protect himself against any excess judgment.” Long, The Law of Liability Insurance § 5A.18, at 5A-107 (2000).
Having paid the policy limits to the Boldens after judgment, it would seethat the LLJUA has no continuing duty to defend Leitrim’s. See Premier InCo. of Mass. v. Furtado, 428 Mass. 507, 511 (1998). Compare Aetna, 3Mass. App. Ct. at 157-158, where the insured requested ongoing defense bthe insurer pursuant to the relevant policy provision. In contrast, the insured ithis case (Leitrim’s) has no desire to have its insurer defend it any further,
Given all this, the LLJUA fails to persuade us that a policy-derived interest *66entitles it to intervene.
In addition, the absence of the insurance policy in the record appendix also precludes our ability to explore the parameters of Leitrim’s duty, as the insured, to cooperate with the LLJUA in the LLJUA’s defense of Leitrim’s. See Guity v. Commerce Ins. Co., 36 Mass. App. Ct. 339, 342 (1994) (insurance policy required insured to cooperate with insurer in investigation, settlement, and defense of any claim or lawsuit). We note, however, that this issue was not raised by the parties.
The evidence likely will focus on insurance industry practices in similar circumstances, expert testimony that the insurer violated sound claims practices, the LLJUA’s field investigation, and advice given to the LLJUA on the probability of success at trial. See Van Dyke, 388 Mass. at 677; Hartford Cas. Ins. Co., 417 Mass. at 119-120; Miller v. Risk Mgmt. Foundation of Harvard Med. Insts., Inc., 36 Mass. App. Ct. 411, 419 (1994).
Moreover, having to engage in subsequent litigation is not, in and of itself, an impairment to a proposed intervener’s interest which justifies intervention. See Motor Club of America Ins. Co., 9 Mass. App. Ct. at 189.
The LLJUA argues that its interest is heightened by the fact that the LL-JUA could be rendered insolvent by a large judgment in the G. L. c. 93A action. However, the record contains no evidence in support of the LLJUA’s allegations. In addition, the LLJUA maintains that the G. L. c. 93A action raises substantial questions concerning the LLJUA’s statutory authority to pay a bad faith judgment. This issue is relevant to the G. L. c. 93A suit, not the present case, and the LLJUA may address this issue if and when a bad faith judgment is rendered against it there.
We need not discuss the final criterion, whether the proposed intervener’s interest is adequately represented by existing parties, for it is evident that the LLJUA’s interest would not be adequately represented in the dramshop case by either the Boldens or Leitrim’s.
As we observed in Cosby, 32 Mass. App. Ct. at 395 n.8, “we are aware *70of no decision in the courts of this Commonwealth that has reversed the denial of a motion to intervene as matter of law under Mass.R.Civ.P. 24(a)(2).” Indeed, it is also rare for our appellate courts to reverse the denial of a rule 24(a)(2) motion under the abuse of discretion standard. In the two instances it has occurred, Cosby, supra, and McDonnell v. Quirk, 22 Mass. App. Ct. 126 (1986), it is fair to say that reasonable minds had little basis for disagreement with the result on appeal.
The LLJUA points to Herman v. Sunshine Chem. Specialties, Inc., 257 N.J. Super. 533 (1992), revd. on other grounds, 133 N.J. 329 (1993), in which the court upheld intervention pursuant to New Jersey Rule 4:33-2, which is similar to Mass.R.Civ.P. 24(b), on facts similar to those in the present case. In Herman, a jury found the defendant liable on the plaintiffs negligence claim and awarded damages. 257 N.J. Super, at 537. The defendant’s insurance company paid part of the judgment, but refused to pay the punitive damages portion of the judgment. Ibid. Subsequently, the defendant declined to appeal the punitive damages portion and assigned its rights against its law firm and insurer to the plaintiff. Ibid. The trial court permitted intervention by the law firm and insurer to prosecute the original defendant’s appeal. Id. at 538. Although the facts are similar to the present case, we do not find Herman to be persuasive authority, as the court engaged in no analysis and merely concluded that the interveners had a sufficient interest in the action, which was not the inquiry directed by the rule. Id. at 538-539.
The LLJUA’s motion to intervene was not accompanied by a pleading setting forth the claim or defense for which intervention was sought, as required by rule 24(c), 365 Mass. 770 (1974). See Board of Selectmen of Stockbridge v. Monument Inn, Inc., 8 Mass. App. Ct. 158, 162 (1979).