Virginia H. Nesbit and the W. Wallace Nesbit Trust (“plaintiffs”) brought this ac tion against Steve McNeil and Black & Company, Inc. (“defendants”) and alleged that the defendants had churned the plaintiffs’ investment accounts. Among other things, plaintiffs sought to recover for violations of the federal securities laws [Securities and Exchange Act of 1934 § 10(b), 15 U.S.C. § 78j(b), and Rule 10b-5, 17 C.F.R. § 240 .10b-5] and under the State of Oregon securities laws [Or.Rev.Stat. § 59.135 (1987)]. The district court directed a verdict against the plaintiffs on the Oregon securities law claim, and submitted the federal securities claim to the jury. The jury brought in a verdict against defendants, and awarded damages in the amount of the excess commissions generated by the churning of the plaintiffs’ accounts. The district court denied a motion for judgment notwithstanding the verdict, and entered judgment accordingly.

Defendants now appeal and claim that the district court erred because it did not permit the offset of trading gains against the excess commissions, because the evidence of churning was insufficient to support the verdict, and because the plaintiffs’ claims were barred by the statute of limitations in whole or in part. Defendants also claim that they should not have been required to disgorge the full amount of excess commissions, but only their net gain on those commissions.

In their cross-appeal, plaintiffs assert that the district court should not have granted a directed verdict on the Oregon securities law claim, since plaintiffs should be able to recover attorney’s fees, even if they were unable to prove any other damages under Oregon law.

We affirm the district court on each of these issues.

BACKGROUND FACTS

Virginia H. Nesbit was a retired school teacher and the widow of W. Wallace Nes-bit, a businessman. Upon his death, Mr. Nesbit left a portfolio of securities that were rather conservative although not necessarily highly successful. Those, as well as other assets, were divided between Mrs. Nesbit and the W. Wallace Nesbit Trust (“the Trust”). Mrs. Nesbit was the trustee of the Trust. From then until 1974, the investments remained conservative and did not do very well. By 1974, there had been a significant loss of value. Mrs. Nesbit then opened accounts for herself and the Trust at Black & Company, Inc. They were opened through Steve McNeil, who was the son of a friend of Mrs. Nesbit. The equity in Mrs. Nesbit’s account was then $167,463, and the equity in the Trust’s account was $44,177. Mrs. Nesbit, who was not knowledgeable in these matters, told the defendants that her investment objectives for herself and the Trust were stability, income and growth. Defendants claim that she told them she wanted to recoup the losses that had been suffered previously.

Defendants then embarked on a course of conduct that extended over a period of eleven and one half years. By the time the accounts were closed out in October of 1985, the equity in Mrs. Nesbit’s account was $301,711, and the equity in the Trust’s account was $92,844. There can be little dispute that this was a substantial increase in value. However, the activities of defendants during those eleven and one half years are called into question in this case.

Plaintiffs have pointed out that defendants first liquidated some of the securities in plaintiffs’ portfolio. Mr. McNeil then embarked on a course of trading that involved 150 issues, one thousand trades, and an overall transaction value of $4,400,000. While the plaintiffs’ account values did grow by $182,915 during the period in question, the defendants’ commissions came to $250,000. Moreover, the investments chosen by defendants were not the kind of investments that one would purchase if one sought a stable, income-producing portfolio. Rather, they were often speculative in nature and were not income- producing. By the time the accounts terminated, many of the investments had accrued losses.

By 1984, Mrs. Nesbit became concerned about the level of activity in the accounts. She kept in closer contact with Mr. McNeil, and the level of trading decreased, but did not end entirely. She became even more concerned in 1985. At that time she discovered losses in the portfolio when calls were made upon her by lenders to whom she had pledged certain of the securities. Her concerns increased when the handling of the accounts was questioned by Ronald Linn, an analyst at Titan Capital, and were not particularly allayed when visits with Mr. McNeil brought forth an apology and an expression of embarrassment at the list of losing stocks. All of this ultimately led to the closing of the accounts in October of 1985. Over one year later, plaintiffs filed this action.

JURISDICTION AND STANDARD OF REVIEW

The district court had jurisdiction over the federal claims pursuant to 28 U.S.C. § 1331 , 15 U.S.C. § 78aa, and 18 U.S.C. § 1965 , and it had pendent jurisdiction over the state claims. We have jurisdiction pursuant to 28 U.S.C. § 1291 .

We review questions of law de novo. United States v. McConney, 728 F.2d 1195, 1201 (9th Cir.) (en banc), cert. denied, 469 U.S. 824 , 105 S.Ct. 101 , 83 L.Ed.2d 46 (1984). That same standard applies to our review of jury instructions, when it is claimed that they are erroneous as a matter of law. Collins v. City of San Diego, 841 F.2d 337, 340 (9th Cir.1988). However, the trial judge has wide latitude in tailoring instructions, and we only review them to assure that they adequately covered the issues presented by the case. Los Angeles Memorial Coliseum Comm’n v. National Football League, 791 F.2d 1356 , 1360 (9th Cir.1986), cert. denied, 484 U.S. 826 , 108 S.Ct. 92 , 98 L.Ed.2d 53 (1987).

We apply the same standard of review to rulings on motions for directed verdicts and rulings on motions for judgment notwithstanding the verdict. As we said in Los Angeles Memorial Coliseum Comm’n v. National Football League, 791 F.2d at 1360:

If reasonable minds could differ about the verdict, neither a directed verdict nor a judgment notwithstanding the verdict is proper. See Peterson v. Kennedy, 771 F.2d 1244, 1256 (9th Cir.1985), cert. denied, 475 U.S. 1122 , 106 S.Ct. 1642 , 90 L.Ed.2d 187 (1986).

DISCUSSION

Although the principal question before us is whether plaintiffs can recover damages for churning when they have had an increase in portfolio values that exceeds the amount of commissions they were charged, the defendants have also claimed that there was insufficient evidence to support the verdict, and that the action is barred by the statute of limitations. If defendants were to prevail on either of the latter issues, there would be no need to consider the damage issue. Therefore, we will address them first.

A. Sufficiency of the Evidence.

The detection and proof of churning is not a simple matter. Churning can only be identified when one considers the whole history of an account, and even then expert testimony is virtually essential. Shad v. Dean Witter Reynolds, Inc., 799 F.2d 525, 530 (9th Cir.1986). As we explained in Mihara v. Dean Witter & Co., Inc., 619 F.2d 814, 820-21 (9th Cir.1980):

Plaintiffs presented substantial evidence on each of these elements, as our statement of background facts has shown. This case involves a relatively unsophisticated investor, who relied upon a person in whom she had confidence — the son of a family friend — to handle her portfolio in a safe and income-generating manner. There can be little doubt that Mr. McNeil did exercise a great deal of de facto control over that account, and the mere fact that he told his client what was being done does not change that situation. This case is quite unlike Brophy v. Redivo, 725 F.2d 1218 (9th Cir.1984), where the only real claim was that certain transactions had been executed without plaintiff’s permission and even against her directions. Here the gravamen of the complaint is that the defendants used their position to overtrade in the account and to do so in a way that they knew did not meet the client’s true investment desires and objectives. See Follansbee v. Davis, Skaggs & Co., Inc., 681 F.2d 673, 676 (9th Cir.1982); Mihara v. Dean Witter & Co., 619 F.2d at 821 ; Hecht v. Harris, Upham & Co., 430 F.2d 1202, 1209 (9th Cir.1970). In other words, there was a good deal of evidence about the relationship between the parties and Mr. McNeil’s control over the account. That is underscored by Mr. McNeil’s embarrassment when he was confronted with the condition of certain securities toward the end of the period. His apology showed that he had something to be embarrassed about; he was not simply executing orders at the direction of Mrs. Nesbit. In addition, the testimony of plaintiffs’ expert showed that one could infer the necessary degree of scienter arising out of the handling of this account, and demonstrated that there was excessive trading.

Defendants contend that there was insufficient evidence of excessive trading in the account. Relying on this circuit’s comments that expert testimony is virtually essential, see Shad, 799 F.2d at 530 , defendants argue that Nesbit’s expert did not present objective statistical evidence of excessive trading. In particular, they complain that the expert did not testify about a turnover ratio for the account as a whole or for an annualized period and that the commission ratio indicated there was no churning.

Although courts often rely on the turnover ratio and commission ratio to indicate excessive trading, no single factor or test identifies excessive trading. See 2 A. Bromberg & L. Lowenfels, Securities Fraud and Commodities Fraud § 5.7 (310) & (322) (1988). Certainly these ratios can make the presence of excessive trading fairly obvious. Id. However, that does not mean that lower ratios will preclude a finding of excessive trading.

Here, Mr. Olson testified that in his expert opinion the 1,000 trades, the high amount of commission compared to the value of the account, the volume of trading, and the presence of losing stock that had been held for some time all pointed toward an improper handling of this account, considering the investment objectives of the client. While defendants have presented evidence and arguments to the contrary, we are in no position to say that the jury improperly found against them on this record. We will not second guess the jury’s determination of the facts, where, as here, the evidence is in conflict and there is substantial evidence to support the jury’s decision.

B. Statute of Limitations.

Defendants claim that the federal securities claims were barred by the statute of limitations.

Defendants’ first attack is upon our decisions in Davis v. Birr, Wilson & Co., Inc., 839 F.2d 1369 (9th Cir.1988) (per curiam), and Volk v. D.A. Davidson & Co., 816 F.2d 1406 (9th Cir.1987). Defendants claim that those decisions are incorrect and that we should revisit them en banc. On October 20, 1989 we rejected the request for an en banc review of our prior decisions, and as a panel we are bound by those decisions. The fact that a concurring opinion in Davis v. Birr, Wilson & Co., Inc., 839 F.2d at 1370 , suggested a different approach is insufficient grounds for this panel to change our longstanding rule. Nor does the fact that the Third Circuit Court of Appeals adopted defendants’ position, in an opinion authored by the same judge who filed the concurring opinion in Davis , permit us to adopt a different approach for this circuit. See In re Data Access Systems Securities Litigation, 843 F.2d 1537 (3d Cir.) (en banc), cert. denied, - U.S. -, 109 S.Ct. 131 , 102 L.Ed.2d 103 (1988).

Moreover, recent Supreme Court cases do not compel a reassessment and reversal of our position. See Agency Holding Corp. v. Malley-Duff & Assocs., Inc., 483 U.S. 143 , 107 S.Ct. 2759 , 97 L.Ed.2d 121 (1987); and Wilson v. Garcia, 471 U.S. 261 , 105 S.Ct. 1938 , 85 L.Ed.2d 254 (1985). We note, in any event, that we have been most reluctant to apply such a determination retroactively in a manner that would cut off the rights of a plaintiff whose action was timely under our decisions which existed at the time the action was filed. See Usher v. City of Los Angeles, 828 F.2d 556, 558-60 (9th Cir.1987).

Defendants assert that even if we use the Oregon statute of limitations of two years, we should find that plaintiffs did not timely file their action. Moreover, defendants claim that at the very least plaintiffs should only recover for transactions that actually took place within two years preceding the filing of this case. Defendants are wrong on both scores.

As this court noted in Shad v. Dean Witter Reynolds, Inc., 799 F.2d at 530 : “Churning is a unified offense which is identified only by an analysis of the entire history of a broker’s management of an account.” In so stating, we relied on Miley v. Oppenheimer & Co., Inc., 637 F.2d 318, 327 (5th Cir.1981), where the court said:

The very fact that churning is a unified offense of some complexity means that a party can hardly be barred on a trade-by-trade basis. Were the law otherwise, by the time an abused client caught on to the wrongdoing, most of the damage may well have been done and the broker would be able to retain ill-gotten gain. In this respect, the situation is analogous to the more usual securities fraud case. As we noted in Volk v. D.A. Davidson & Co., 816 F.2d at 1412 : “In fraud cases, a cause of action is generally said to accrue when a defendant commits the last overt injurious act_ However, the statute of limitations is not triggered until the defrauded individual has actual or inquiry notice that a fraudulent misrepresentation has been made.” This is not to say that a client can sleep on her rights interminably. Here some evidence could have supported a decision that plaintiff waited too long. It could be argued that the last act of churning took place before Mrs. Nesbit became concerned in 1984, and that her knowledge in 1984 was sufficient to place her on at least inquiry notice. That is, it could be asserted that at the time of the slowdown of the trading in the accounts, or perhaps at some earlier date, she should have done more. However, the evidence of plaintiffs’ knowledge and of the defendants’ activities and representations was in conflict. For exam- pie, Mr. McNeil continued to encourage her to purchase securities and assured her that the losing stocks would recover. It was proper for the trial court to submit those issues to the jury on the statute of limitations questions. The jury found that the statute of limitations did not bar this action. We are unable to say that its determination was so devoid of evidentiary support as to require reversal. In other words, it was proper to find that the earliest the statute of limitations began to run was at the time the final transactions in the account were consummated, because churning activity did not actually end until then. That was in 1985. The whole complex history of the account could then be scrutinized, and one could then determine whether the plaintiffs had been wronged, even though that wrong may have been hidden by the apparent overall success of the investment strategy adopted by the defendants. We need not decide when plaintiffs may finally have been placed on inquiry notice, since it was reasonable for the jury to find that the last overt fraudulent act took place within the statute of limitations period.

Defendants, therefore, cannot succeed by asserting the statute of limitations.

C. The Measurement of Damages.

As we have already noted, defendants obtained commissions of $250,000 from the plaintiffs. The jury found that $134,000 of that constituted excess commissions. At the same time, the value of plaintiffs’ accounts increased in the sum of $182,915. Defendants claimed below, and continue to claim, that the plaintiffs’ portfolio gain should be offset against the plaintiffs’ commission loss, as a result of which plaintiffs can recover no damages whatever. The district court disagreed, and gave the following instruction to the jury: “If you find that the plaintiffs have proven their claims for churning, excessive trading, plaintiffs may recover as damages any commissions they paid as a result of the churning in excess of commissions that would have been reasonable on transactions during the pertinent time period.” The district court did not go on to instruct the jury that it could then offset the trading gains against those commission losses. We agree with the district court.

We begin with the rather straightforward principle announced in Mihara v. Dean Witter & Co., Inc., 619 F.2d at 826 , where we said that, “While damage for churning are limited to commissions and interest, plaintiff’s claim as to the suitability of the securities purchased would also encompass trading losses.” As the Fifth Circuit Court of Appeals explained in Miley v. Oppenheimer & Co., Inc., 637 F.2d at 326 , there are two separate and distinct possible harms when an account has been churned, and those are:

In the case at hand, the plaintiffs only suffered one of those harms, but there is no reason to find that they should be denied a recovery because their portfolio increased in value, either because of or in spite of the activities of the defendants.

We are mindful of our decision in Hatrock v. Edward D. Jones & Co., 750 F.2d 767 (9th Cir.1984). In that case we pointed out that the investor could recover both the “excessive commissions charged by the broker, and the decline in value of the investor's portfolio resulting from the broker’s fraudulent transactions.” Id. at 774 . However, the investor had in fact suffered both of those losses, and nothing in that opinion suggested that we intended to declare that an investor must suffer both harms before he can recover from the broker. It is not surprising that a broker who is guilty of churning — an element of which is the making of unsuitable investment decisions — may also have purchased unsuitable investments. That does not mean that the two concepts should be conflated, and we have never said that they must be. Rather, our decisions are properly harmonized to reach an unremarkable result which ultimately makes the plaintiff whole in any event — a plaintiff may separately recover either or both types of damages, but gains in portfolio will not offset losses in commissions. That is a result that prevents the broker from escaping with improper commissions. Defendants would have us say that a broker who, for example, engages in unsuitable management of the account but actually buys suitable or successful investments cannot be held responsible for his actions. One can easily envision a single successful security which is bought and sold an unreasonable number of times, thereby relieving the client of gains and improperly enriching the broker. We would be remiss if we were to find no redress within the securities law for that kind of wrongdoing. Finally, it would be a poor rule indeed that unnecessarily forced courts and juries to enter into the complex determinations needed to determine whether any gain on a portfolio was due to the improper activities of a broker. Once the trier of fact finds that the trading was excessive it has determined that the trading should not have taken place. It should not be forced to decide whether gains or losses are a result of market forces, luck, good times, or intrinsically good stock. Nor should the trier of fact have to decide what part of the loss or gain should be allocated to any given factor. That is one reason we have said that proof of causation is not a necessary element of a plaintiffs case. See Hatrock v. Edward D. Jones & Co., 750 F.2d at 773 . The broker will have made trades that should not have been made, and he should not retain his ill gotten gain. This salutory rule also informs the brokerage community that churning is a fraud that will violate the securities laws, regardless of the ultimate condition of the client’s portfolio. It therefore serves to forward the deterrent policies which underlie the federal securities laws. Randall v. Loftsgaarden, 478 U.S. 647, 664 , 106 S.Ct. 3143, 3153 , 92 L.Ed.2d 525 (1986).

As a final attack on damages, the defendants rely on a portion of the Supreme Court’s decision in Randall v. Loftsgaar-den, id., for the proposition that the plaintiff must lean upon the concept of unjust enrichment in order to recover. Reasoning from that proposition, defendants ask that the brokers’ expenses be deducted from the commissions and that plaintiffs recover, at most, the difference. It is true that Randall addresses the question of unjust enrichment at 478 U.S. 661 -64, 106 S.Ct. 3152 -53, but it does not mandate the theory suggested by defendants. The core of the defendants’ position must be that if a broker defrauds a client of a commission, the broker ought to be able to deduct the expenses he incurred in perpetrating the fraud and then return the rest to the client. We believe that the very statement of the proposition refutes it, and not a shard of judicial precedent supports it. Randall , on the other hand, merely offers the possibility that even if a plaintiff has not lost as much as a defendant has gained, the plaintiff should recover that gain from the defendant. To apply that to a situation where defendant has not netted as much as he has taken from the plaintiff would stand Randall on its head. It- takes no hiero-phant to discover that.

Therefore, we must reject defendants’ assault on the damage award in this case.

D. Plaintiffs’ Claim under Oregon Securities Law.

The district court determined that plaintiffs had not suffered an injury under Or.Rev.Stat. § 59.115(2) (1987), which allows recovery of damages and fees by a buyer of securities from a seller who violates Or.Rev.Stat. § 59.135 (1987). On appeal plaintiffs ask us to enter a directed verdict for them on that claim and to award them attorney’s fees. We are unable to do that.

Even assuming that the Oregon legislature intended that churning would violate section 59.135, plaintiffs may not rely on section 59.115(2) for an award of either damages or fees. That section by its language applies only to a buyer-seller relationship. Here, the plaintiffs and defendants were in a principal-agent relationship.

Plaintiffs also contend that even if section 59.115(2) would not provide them with damages and attorney’s fees, we should imply a remedy for violations of section 59.135 because such a remedy exists under federal law. While the Oregon Supreme Court has not spoken to this precise issue, we do note that it has resisted attempts to imply rights to damages under section 59.-135. Held v. Product Mfg. Co., 286 Or. 67 , 592 P.2d 1005 (1979). We recognize that section 59.135 is patterned after Rule 10b-5. Shivers v. Amerco, 670 F.2d 826, 831 (9th Cir.1982). That does not affect our decision. In fact, as Held v. Product Mfg. Co., 286 Or. 67 , 592 P.2d at 1007 , shows, the Oregon Supreme Court’s reticence about expanding remedies is so strong that it has refused to imply a remedy for sellers of securities when the statutory scheme only provided an express remedy for buyers in section 59.115(2). We therefore doubt that it would imply a remedy against brokers. Even if it did, there is no reason to think that it would then go on to imply an attorney’s fee recovery for the violation, since even federal law does not allow recovery of fees.

To put it succinctly, plaintiffs’ chain of reasoning is too frangible to bear the weight plaintiffs seek to hang upon it.

CONCLUSION

We have been presented with the rather unusual case of plaintiffs whose portfolios increased while under the guidance and control of the defendants, and who still chose to bring an action to recover commissions that they had paid to those same defendants.

The jury before which this case was tried could have decided that defendants were perfectly honest brokers, who were being victimized by rather greedy clients. The jury did not do so. Instead, it found that plaintiffs were indeed wronged by the defendants’ churning of the accounts. That determination was supported by substantial evidence. As we have noted, the jury’s determination that the statute of limitations had not run in this case was also properly supported.

As a result, the defendants were properly required to disgorge the inappropriate portion of their commissions, even if the portfolio itself increased in value, since, as we have shown, issues regarding the performance of the portfolio are separate from issues related to excess commissions. Just as there can be gains or losses when trading is appropriate, there can be gains or losses when it is inappropriate. The propriety of the trading determines the right to a commission. The same evidence may also influence the right to recover for losses, but that does not mean that gains should undermine the plaintiffs’ right to recover improper commissions. The wrongs are separate, and the result of each should be analyzed separately.

However, plaintiffs do attempt to reach too far when they ask this court to append an attorney’s fee remedy to their recovery by stretching and bending Oregon law for that purpose.

Thus, the district court’s rulings were correct.

AFFIRMED.

. We use here plaintiffs’ estimates. Defendants’ would raise these to $338,615 and $106,463 respectively. . Using defendants’ numbers from footnote 1 this could be $233,438. . The turnover ratio is the ratio of the total cost of the purchases made for the account during a given period of time to the amount invested. 2 A. Bromberg & L. Lowenfels, Securities Fraud and Commodities Fraud § 5.7(322) (1988). . The commission ratio is "the ratio of the broker's commissions generated by the account to the size of the customer’s investment in that account.” 2 A. Bromberg & L. Lowenfels, Securities Fraud and Commodities Fraud § 5.7(322) (1988). . The plaintiffs claim that defendants did not preserve this contention below. That is not so. Defendants vigorously presented their theory at every turn, and the district court specifically indicated that objections to its instructions were preserved. While the defendants did accept the instructions given as being a sufficient statement of the law as the district court saw it, they never agreed that the district court saw the law properly. . Appellants insist that this decision is contrary to our decision in Arrington v. Merrill, Lynch, Pierce, Fenner & Smith, Inc., 651 F.2d 615 (9th Cir.1981). We disagree. In Arrington , the issue was portfolio loss, and, of course, dividends gained on the portfolio were properly considered in that context. Here, as we have said, we focus on a loss that would occur whether the portfolio value went up, down, or stayed the same. It arose out of commissions paid for trading that never should have taken place. That, as we have pointed out, is a separate wrong for which damages are separately recoverable.