There can be no fraud if the bargain be merely a fair contest, or trial of judgment. In all contracts, each party naturally and fairly attempts to obtain advantage. As in a contract of sale, the vendor endeavours to extol the article, the vendee to depreciate; each exercises his own judgment, and neither party can be said to be guilty of a fraud in making bare assertions, upon which the other party probably places no reliance, and which he does not embody in his contract.
A practical treatise on the law of contracts, not under seal; and upon the usual defences to actions thereon. Joseph Chitty, 1834.
In Sanders v McLean (1883) 11 QBD 327 at p.343 per Lord Justice Bowen:
A long line of authority establishes that if an insured makes a fraudulently inflated claim under the policy he forfeits any lesser claim which he could properly have made. An owner who claims $ 10,000,000, knowing that the claim could not possibly be worth more than $ 9,000,000 recovers nothing. The rule, which was established by the early nineteenth century, was described by Mance LJ in Axa General Insurance v Gottlieb  Lloyd’s Rep IR 369 as a special common law rule. It applies even if there is no clause in the policy incorporating it and is designedly draconian. It functions as a deterrent to the deception of insurers who, in the nature of things, will have no, or very little, knowledge of the incident which is said to give rise to the claim. Part of the rationale is that if lying to the insurers did not attract that sanction, the dishonest insured would enjoy a one way bet. In the example given, if the lie was never found out, the insurers might end up paying out $ 10,000,000. If the insured was found out there would be no effective sanction (at any rate where the claim did not come to litigation where costs penalties might apply). He would still recover the true value of his claim.
Per Mance LJ in The Aegeon  EWCA Civ 247 at paras 36-38 and 45:
36. What relationship need there then be between any fraud and the claim if the fraudulent claim rule is to apply? And need the fraud have any effect on insurers' conduct? Speaking here of a claim for a loss known to be non-existent or exaggerated, the answers seem clear. Nothing further is necessary. The application of the rule flows from the fact that a fraudulent claim of this nature has been made. Whether insurers are misled or not is in this context beside the point. The principle only arises for consideration where they have not been misled into paying or settling the claim, and its application could not sensibly depend upon proof that they were temporarily misled. The only further requirement is that the part of the claim which is non-existent or exaggerated should not itself be immaterial or unsubstantial…
37. What is the position where there is use of a fraudulent device designed to promote a claim? I would see no reason for requiring proof of actual inducement here, any more than there is in the context of a fraudulent claim for non-existent or exaggerated loss. As to any further requirement of "materiality", if one were to adopt in this context the test identified in the Royal Boskalis case  LRLR 523 and The Mercandian Continent  2 Lloyd’s Rep 563, then, as I have said, the effect is, in most cases, tantamount to saying that the use of a fraudulent device carries no sanction. It is irrelevant (unless it succeeds, which only the insured will then know). … what is material at the claims stage depends on the facts then known and the strengths and weaknesses of the case as they may then appear. It seems irrelevant to measure materiality against what may be known at some future date, after a trial. The object of a lie is to deceive. The deceit may never be discovered. The case may then be fought on a false premise, or the lie may lead to a favourable settlement before trial. Does the fact that the lie happens to be detected or unravelled before a settlement or during a trial make it immaterial at the time when it was told? In my opinion, not. Materiality should take into account the different appreciation of the prospects which a lie is usually intended to induce on insurers' side, and the different understanding of the facts which it is intended to induce on the part of a judge at trial.
38. The view could, in this situation, be taken that, where fraudulent devices or means have been used to promote a claim, that by itself is sufficient to justify the application of the sanction of forfeiture. The insured’s own perception of the value of the lie would suffice. Probably, however, some limited objective element is also required. The requirement, where a claim includes a non-existent or exaggerated element of loss, that that element must be not immaterial, "unsubstantial" or insignificant in itself offers a parallel. In the context of use of a fraudulent device or means, one can contemplate the possibility of an obviously irrelevant lie—one which, whatever the insured may have thought, could not sensibly have had any significant impact on any insurer or judge. Tentatively, I would suggest that the courts should only apply the fraudulent claim rule to the use of fraudulent devices or means which would, if believed, have tended, objectively but prior to any final determination at trial of the parties' rights, to yield a not insignificant improvement in the insured’s prospects—whether they be prospects of obtaining a settlement, or a better settlement, or of winning at trial. Courts are used enough to considering prospects, e g when assessing damages for failure by a solicitor to issue a claim form within a limitation period. …
45. What then is the appropriate approach for the law to adopt in relation to the use of a fraudulent device to promote a claim, which may (or may not) prove at trial to be otherwise good, but in relation to which the insured feels it expedient to tell lies to improve his prospects of a settlement or at trial? The common law rule relating to cases of no or exaggerated loss arises from a perception of appropriate policy and jurisprudence on the part of our 19th century predecessors, which time has done nothing to alter. The proper approach to the use of fraudulent devices or means is much freer from authority. It is, as a result, our duty to form our own perception of the proper ambit or any extension of the common law rule. In the present imperfect state of the law, fettered as it is by section 17, my tentative view of an acceptable solution would be: (a) to recognise that the fraudulent claim rule applies as much to the fraudulent maintenance of an initially honest claim as to a claim which the insured knows from the outset to be exaggerated; (b) to treat the use of a fraudulent device as a sub-species of making a fraudulent claim—at least as regards forfeiture of the claim itself in relation to which the fraudulent device or means is used (the fraudulent claim rule may have a prospective aspect in respect of future, and perhaps current, claims, but it is unnecessary to consider that aspect or its application to cases of use of fraudulent devices); (c) to treat as relevant for this purpose any lie, directly related to the claim to which the fraudulent device relates, which is intended to improve the insured’s prospects of obtaining a settlement or winning the case, and which would, if believed, tend, objectively, prior to any final determination at trial of the parties' rights, to yield a not insignificant improvement in the insured’s prospects—whether they be prospects of obtaining a settlement, or a better settlement, or of winning at trial; and (d) to treat the common law rules governing the making of a fraudulent claim (including the use of fraudulent device) as falling outside the scope of section 17… On this basis no question of avoidance ab initio would arise.
Per Clarke LJ in Versloot Dredging BV & Anor v HDI Gerling Industrie Versicherung AG & Ors  EWCA Civ 1349 at paras 109, 112:
109. The foundation of the [fraudulent devices] rule is the obligation of the utmost good faith - an incident of the special relationship between insured and insurer. The effect of the rule is that if you lie to your insurer in respect of anything significant in the presentation of the claim you will not recover anything from him. Once it is accepted, as it has long been, that an assured who fraudulently exaggerates his claim forfeits the whole of it - so that a fraudulent claimant can lose almost all of his otherwise valid claim (say 95%) - there seems to me no satisfactory reason in principle why the user of a fraudulent device should not lose 100%. It is true that the user of such a device may turn out not to have been seeking a benefit to which he was not entitled; but the same is true, in the example given, in respect of the fraudulent claimant as regards most of his claim. In the case of both the fraudulent claim and the fraudulent device, fraud is used to obtain something to which the insured is either not entitled or would not otherwise have received: in a fraudulent claim case, the bogus part of the claim; and in a fraudulent device claim, earlier payment than full investigation would otherwise permit.
112. The harshness of the result of a fraudulent devices rule is most apparent when, as here, the Court has, in the end, determined that the claim is otherwise valid. But the rule is directed to an earlier stage. Fraudulent devices, as their definition shows, are used by those who think their use desirable in order to bolster a claim which appears to have potential weaknesses (e.g. as here Owners' possible responsibility for a deficiency in the bilge alarm system) and to avoid or cut short lines of inquiry or investigation which might prevent or postpone the payment of it. At the stage when they are used it will probably not be possible to tell whether the claim is one that will in the end be accepted, or held, to be valid. The risk to the insurer is that the device may achieve its purpose, so that the insurer fails to explore the claim properly and pays out in respect of a claim where he may have a defence. If he does do so it will never be known whether, had he not done so, the result would have been the same.