David Tachau (argued and briefed), Maureen Taylor, Brown, Todd & Heyburn, Louisville, Ky. and Paul E. Sullivan, Brown, Todd & Heyburn, Lexington, Ky., for plaintiff-appellant.

Karen J. Greenwell, William H. McCann, Wyatt, Tarrant & Combs, Lexington, Ky., Earle J. Schwarz (argued and briefed), Frank L. Watson, Saul C. Belz, and Waring Cox, Memphis, Tenn., for defendants-appellees.

Before: GUY and BOGGS, Circuit Judges; and HARVEY, Senior District Judge.*

JAMES HARVEY, Senior District Judge.

This case arises out of the Bank of Lexington & Trust Company's (the Bank's) purchase of municipal bonds from Vining-Sparks Securities, Inc. (Vining-Sparks), a brokerage firm. When the Bank began to suffer losses on the bonds, it brought suit against Vining-Sparks and James Vining, the president and CEO of Vining-Sparks, alleging violations of the Securities Exchange Act of 1934, 15 U.S.C. § 78j(b), the Racketeer Influenced and Corrupt Organizations Act (RICO), 18 U.S.C. § 1962, Kentucky's "Blue Sky" law, Ky.Rev.Stat. § 292.320, and Kentucky common law. The Bank sought rescission of the sales contract and damages, contending that Vining and Vining-Sparks made material misrepresentations, failed to disclose material information, excessively marked-up the bonds' prices, and breached both contractual and fiduciary duties.

After a six-week bench trial, the district court ruled in favor of Vining and Vining-Sparks, finding, in short, that the Bank understood the nature of the bonds before buying them, yet bought them because of certain anticipated tax advantages. The Bank appeals. Rejecting each of the Bank's grounds for appeal, we AFFIRM.

Between July and December 1985, the Bank bought from Vining-Sparks in fourteen separate transactions zero-coupon bonds from ten different bond issues. Zero-coupon bonds are debt securities on which no interest is paid. When the bonds mature, the holder collects a single payment. Accordingly, investors buy the bonds at a deep discount from face value.

William Lane, a Vining-Sparks account executive, had encouraged the Bank's investment officer, Ann Hall, to buy the bonds because of their tax benefits. On July 9, 1985, several months after Lane began discussing with Hall the merits of zero-coupon bonds, Lane and Vining met with Hall, James Rose, the Bank's principal owner, Clyde Mauldin, the Bank's president, and John Sullivan, an accountant employed by Rose, to talk about the Bank's possible purchase of zero-coupon bonds. Vining presented Vining-Sparks's research into zero-coupon bonds, detailing the tax benefits Vining-Sparks believed were available, and distributing written materials which included projected yields for various bond issues. He explained that Vining-Sparks had discovered that investors could treat zero-coupon municipal bonds issued before September 3, 1982 differently for tax purposes than for accounting purposes. He asserted that by accreting the bonds on a straight-line basis for tax purposes, while accreting them on a constant-yield basis for accounting purposes, an investor could later sell the bonds at a tax loss, yet not necessarily at an accounting loss. The investor could then use the tax loss to shelter ordinary income.

Following the meeting, Sullivan sent Vining-Sparks's research report to the Bank's outside auditors, Peat, Marwick, Mitchell & Company (Peat Marwick) for review. Sullivan asked Peat Marwick to confirm "that if a bank were involved in a 'rolling' type of investment in these bonds, then by virtue of the artificial tax losses (in comparison to book losses) significant yields may be sustained on these type of investments." Peat Marwick responded by telephone call that Vining-Sparks had correctly analyzed the tax and accounting treatment of zero-coupon municipal bonds. A later written response concluded that, where interest rates remain constant, an investor could buy and sell such bonds with little impact on book income, but with substantial tax losses.

Soon after receiving Peat Marwick's verbal confirmation, Hall told Lane that the Bank had decided to spend approximately $3 million on zero-coupon bonds. By December 1985, the Bank had spent almost $3.3 million on zero-coupon bonds. Following each sale, Vining-Sparks sent to the Bank a transaction confirmation statement, documenting the sale and describing the bonds sold. Further, for many of the bonds sold, Vining-Sparks sent accretion tables showing projected monthly increases in the value of the bonds.

In 1986, interest rates began to fall. The sudden drop in interest rates quickly impacted the market value of the Bank's zero-coupon bonds. Three of the bond issues from which the Bank had bought zero-coupon bonds were single-family mortgage revenue bonds (SFMRBs). Because of the drop in interest rates, the mortgages financed with revenue generated by the sale of SFMRBs began to be prepaid earlier than expected. Of the remaining bond issues from which the Bank had purchased zero-coupon bonds, five were multi-family mortgage revenue bonds (MFRBs), and two were non-housing revenue bonds. Although a number of factors apparently contributed to the decline in market value of the Bank's MFRBs, it is certain that, as interest rates fell, the aggregate market value of the Bank's zero-coupon bonds began to drop sharply, plunging almost 41 percent by the end of June 1986.

On July 1, 1987, the issuer of one lot of the Bank's SFMRBs, the Olathe, Kansas SFMRBs, recalled $2 million of bonds, resulting in a substantial loss to the Bank. No other issuer of bonds prematurely redeemed zero coupon bonds held by the Bank before the Bank filed suit against Vining and Vining-Sparks on February 5, 1988.

The Bank's principal argument on appeal is that several of the district court's crucial findings of fact are clearly erroneous. Under Rule 52(a) of the Federal Rules of Civil Procedure, "[f]indings of fact, whether based on oral or documentary evidence, shall not be set aside unless clearly erroneous, and due regard shall be given to the opportunity of the trial court to judge of the credibility of the witnesses." As the Supreme Court has stated, " '[a] finding of fact is "clearly erroneous" when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.' " Anderson v. Bessemer City, 470 U.S. 564, 573, 105 S. Ct. 1504, 1511, 84 L. Ed. 2d 518 (1985) (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S. Ct. 525, 542, 92 L. Ed. 746 (1948)). Where the district court's view of the evidence is plausible, we may not reverse its findings of fact even if we would have weighed the evidence differently. Id. at 573-74, 105 S.Ct. at 1511.

Further, although the Bank vigorously attacks the district court for relying heavily in its findings of fact and conclusions of law on memoranda drafted by counsel for Vining and Vining-Sparks, the Court notes that "even when the trial judge adopts proposed findings verbatim, the findings are those of the court and may be reversed only if clearly erroneous." Id. at 572, 105 S.Ct. at 1511.

The Bank sought redress under many theories of liability, but central to several of its theories under the Securities Exchange Act of 1934, 15 U.S.C. § 78j(b), Kentucky's "Blue Sky" law, Ky.Rev.Stat. § 292.320, and Kentucky common law was the assertion that Vining and Vining-Sparks had misrepresented the quality and extent of the firm's research into the bonds. But the district court found no merit to the Bank's assertion, finding instead that Vining-Sparks conducted "deliberate and studied" research.

Both federal and Kentucky securities laws make it unlawful for any person to make any untrue statement of a material fact in connection with an offer for the sale of securities. 17 C.F.R. § 240.10b-5 (Rule 10b-5) (promulgated under 15 U.S.C. § 78j); Ky.Rev.Stat. § 292.320(1)(b). The Bank alleged that Vining-Sparks lied when it said that it would only sell the Bank securities with particular protections against premature redemptions.

The Bank asserted that Vining orally promised that zero-coupon bonds would represent only between 3 and 5 percent of each issue of SFMRBs recommended for purchase, and that the issuer would call those bonds last if the mortgages financed with revenue generated by the sale of SFMRBs began to be prepaid earlier than expected. Yet, the zero-coupon bonds bought by the Bank from each issue of SFMRBs represented considerably more than 3 to 5 percent of the issue. Hence, the Bank argues that Vining misrepresented the research his firm had conducted, or would conduct, before recommending bonds for purchase.

In response to the Bank's allegation, however, Vining testified that he had not promised that the zero-coupon bonds would represent only 3 to 5 percent of each issue of SFMRBs. Rather, he claimed that he had merely said during his presentation to the Bank that zero-coupon bonds could represent as low as 3 to 5 percent of each issue of SFMRBs. Because the district court had the opportunity to view Vining's demeanor when he explained why the Bank might have believed that he had made such a promise, and the demeanor of the witnesses testifying to the contrary, we see no reason to set aside the district court's finding. See Anderson, 470 U.S. at 575, 105 S.Ct. at 1512 ("When findings are based on determinations regarding the credibility of witnesses, Rule 52(a) demands even greater deference to the trial court's findings").

The Bank also challenges the district court's finding that Vining and Vining-Sparks knew the mortgage rates underlying the zero-coupon bonds sold to the Bank, and knew the speeds at which comparable mortgages were being prepaid in 1985. This finding led to the district court's finding that Vining and Vining-Sparks reasonably believed that the bonds sold to the Bank from each bond issue would not be recalled during the holding period contemplated by the Bank.

In support of its challenge to the district court's finding, the Bank points to Vining's deposition testimony:

Q. How would an investor determine that downside risk of early redemption? I mean, if I bought that security ... how would I sit down and compute what happens to me if mortgage interest rates decline?

A. The research that's available today is better than it was in '85.... I think that all the massive fall in interest rates and all the refinancing came after 1985. Prior to then there really hadn't been any work done by anyone on forecasting that I have seen, not the first thing I have seen in terms of prepayment experience.

The Bank then contrasts this testimony with Vining's trial testimony:

Q. Mr. Vining, with respect to other research in connection with the ... zero-coupon bonds, what else did you do? Did you do anything else with respect to speed calculation and forecasting and expectations of prepayment rates?

A. We have done the forecasting of speeds based on current data continuously since 1983.

After asserting that the Bank's losses show that Vining did not conduct the research he claimed at trial to have conducted, the Bank concludes that the district court's findings regarding Vining and Vining-Sparks's research into prepayment speeds are clearly erroneous.

By focusing sharply on single answers in Vining's deposition and trial testimony, the Bank obscures the remainder of his testimony. Although Vining testified in his deposition that he had not forecasted the rates at which mortgages underlying the bonds would be prepaid, he did so only after citing current research methods. At trial, he explained in detail the research he had conducted in determining that the bonds sold to the Bank would have particular protections against early redemptions. With regard to forecasting mortgage prepayment speeds, he noted that after analyzing historical prepayment patterns, he and Vining-Sparks believed that the issuer would not call the bonds before 1995. To show that others shared this belief, he and Vining-Sparks called a housing analyst to testify that he knew of no analyst who had predicted the sudden and deep decline in interest rates, which had led to a surge in mortgage refinancing. Accordingly, the district court's account of the evidence is plausible. Although Vining had not calculated mortgage prepayment speeds in the same manner as currently calculated, he knew the speeds at which mortgages comparable to the mortgages underlying the SFMRBs were being prepaid in 1985.

Finally, the Bank points to three articles which appeared in The Bond Buyer, an industry publication which Vining-Sparks's bond trader, Willie Hickerson, testified he read every day. Each of the articles warns of the risk of early redemptions of SFMRBs caused by unused funds calls. Where bond proceeds are not used to finance mortgages within a specified time period, the issuer must call the bonds. The Bank argues that the district court's finding that Vining-Sparks adequately researched the bonds sold is clearly erroneous because careful research would surely have uncovered the articles and their warnings.

Although one of the articles predicts falling interest rates, and one mentions mortgagor refinancings as a factor affecting early redemptions of mortgage bonds, each article is primarily a warning to investors in SFMRBs about unused funds calls. But Vining-Sparks did not sell the Bank any bonds where the risk of an unused funds call remained. Indeed, Vining and Hickerson each testified that Vining-Sparks had refused to trade bonds in danger of an unused funds call. Thus, the articles do not contain the clear and dire warnings about the perils of mortgage prepayments which the Bank asserts they do.

In sum, the Court is not left with the conviction that the district court committed a mistake when it found that Vining-Sparks had conducted "deliberate and studied" research before recommending bonds for purchase.

The Bank also sought to hold Vining-Sparks liable under the Securities Exchange Act of 1934, 15 U.S.C. § 78j(b), and its implementing regulation, Rule 10b-5, for failing to disclose adequately the call features of bonds with optional call provisions in transaction confirmation statements sent to the Bank following each sale of those bonds. In addition, the Bank alleged that Vining-Sparks violated Rule 10b-5 when it stated in its confirmation statement following the sale of the Olathe, Kansas SFMRBs that the bonds were "non-callable," even though they were subject to prepayment redemption. Rejecting the Bank's allegations, the district court found that Vining-Sparks had furnished the Bank sufficient information about the call features of each group of bonds bought by the Bank. The Bank disputes this finding.

The Bank asserted that Vining-Sparks omitted material information in violation of Rule 10b-5 when it failed to include a statement of yield to call in the confirmation statement for each bond issue with optional call provisions. In support of its assertion that a statement of yield to call is material information, the Bank cited Rule G-15 of the Municipal Securities Rulemaking Board (MSRB), which details the information securities brokers must include in written transaction confirmation statements. The Securities Exchange Act of 1934, as amended, requires brokers to comply with the MSRB's rules. 15 U.S.C. § 78o-4(c)(1). Rule G-15(a)(E) requires brokers to include a "description of the securities, including at a minimum the name of the issuer, interest rate, maturity date and if the securities are ... subject to redemption prior to maturity (callable) ... an indication to such effect...." MSRB Manual--General Rules, G-15(a)(E) (CCH) p 3571 (1985). Further, Rule G-15(a)(i)(I)(2) instructs that, "for transactions effected on the basis of dollar price, the dollar price at which transaction was effected, and the lowest of the resulting yield to call, yield to par option, or yield to maturity shall be shown." Id. Because Vining-Sparks included in the confirmation statements sent to the Bank the yield to maturity, rather than the lower yield to call, the Bank argued that Vining-Sparks had omitted material information.

In response, Vining-Sparks argued that a statement of yield to call was not a fact material to the Bank's decision to buy bonds with optional call provisions, because the Bank did not plan to hold the bonds until their call dates. We need not address that question, however, for even if we assume that yield to call was a material fact, sufficient evidence supports the district court's finding that Vining-Sparks adequately disclosed the call features of each lot of bonds sold. To determine whether a securities fraud defendant adequately disclosed a material fact under Rule 10b-5, the trier of fact must decide whether a reasonable investor would have been misled by the information disclosed. Isquith v. Middle South Utils., Inc., 847 F.2d 186, 207, 211 (5th Cir.), cert. denied, 488 U.S. 926, 109 S. Ct. 310, 102 L. Ed. 2d 329 (1988); Durning v. First Boston Corp., 815 F.2d 1265, 1268 (9th Cir.), cert. denied, 484 U.S. 944, 108 S. Ct. 330, 98 L. Ed. 2d 358 (1987). Further, in reaching its finding, the trier of fact must consider all of the information disclosed, not merely the information alleged to be inadequate. Isquith, 847 F.2d at 207.

In finding that the information Vining-Sparks disclosed would not have misled a reasonable investor, the district court noted that in each of the eight transactions involving bonds with optional call features, the confirmation statement sent to the Bank indicated that the bonds were callable, and set out both the call date and call price. Further, the district court found that shortly after six of the eight transactions, Vining-Sparks furnished the Bank with accretion schedules which included a statement of yield to call as required by MSRB Rule G-15(a)(i)(I)(2). After noting that Hall testified that she reviewed the schedules to verify her understanding of the bonds the Bank had bought, the district court turned to Vining's testimony. He testified that he chose to provide the call price and call date, rather than yield to call, because he believed that the Bank could more easily calculate its after-tax yield using the call price and call date than it could using yield to call. Moreover, he noted that under the proposed tax plan Vining-Sparks had advocated, he did not expect that the Bank would hold the bonds until called. Finally, the district court cited the results of an investigation conducted by the National Association of Securities Dealers, Inc. (NASD), after it received a complaint from the Bank. The NASD is a self-regulatory organization subject to extensive oversight, supervision, and control by the Securities and Exchange Commission (SEC) on an ongoing basis. See 15 U.S.C. § 78s. Although the NASD issued a letter cautioning Vining-Sparks about future technical violations of MSRB Rule G-15, the NASD concluded that the complaint filed by the Bank required no disciplinary action.

After reviewing the district court's account of the evidence in light of the entire record, we find ample support for its finding that Vining-Sparks adequately disclosed the call features of the bonds with optional call provisions. In reaching its finding, the district court properly considered the information disclosed as a whole, see Isquith, 847 F.2d at 207, just as the NASD apparently did. After noting that Vining-Sparks's confirmation statements showed that the bonds with optional call provisions were callable, and disclosed both the call price and call date, information from which yield to call can easily be calculated according to Vining, the district court concluded that a reasonable investor would not have been misled about the bonds' call features. This interpretation of the evidence is not illogical or implausible. Accordingly, we may not reverse the district court's finding. See Anderson, 470 U.S. at 573-74, 105 S. Ct. at 1511.

The Bank also challenges the district court's finding that Vining-Sparks adequately disclosed the call features of the Olathe, Kansas SFMRBs. The Bank points to the confirmation statement, which stated that the bonds were "non-callable," and argues that the district court simply ignored this violation of MSRB Rule G-15 in reaching its finding. A reasonable investor, argues the Bank, would not have foreseen that the issuer could call the bonds if the underlying mortgages were prepaid.

The Bank's argument lacks merit. As noted, in deciding whether a securities broker adequately disclosed a material fact, the trier of fact must reckon with the entire disclosure, not merely those fragments to which the investor points. Hence, as Vining-Sparks notes, the confirmation statement also stated that the bonds were "subject to special redemption." Moreover, the Bank asserted at trial that Vining-Sparks promised to protect the Bank from early redemptions by only selling SFMRBs which the issuer would call last if the mortgages were prepaid earlier than expected. This assertion undermines the Bank's attack on the district court's finding that Vining-Sparks orally disclosed that the issuer could recall the bonds if the underlying mortgages were prepaid. Thus, we are not convinced that the district court made a mistake in finding Vining-Sparks's disclosure adequate.

Finally, the Bank quarrels with the district court's finding that the Bank failed to prove that Vining-Sparks charged an unfair price for several lots of bonds. The Bank sought to recover damages under the Securities Exchange Act of 1934, 15 U.S.C. § 78j(b), and Rule 10b-5, on the theory that Vining-Sparks had committed securities fraud by failing to disclose excessive commissions. While the failure to disclose exorbitant mark-ups violates section 78j(b) and Rule 10b-5, Hoxworth v. Blinder, Robinson & Co., 903 F.2d 186, 200 (3rd Cir.1990); Zero-Coupon Securities, Exchange Act Release No. 34-24368, Fed.Sec.L.Rep. (CCH) p 22,768, at 16,631-24 to -26 (April 21, 1987) [hereinafter SEC Release], the Bank does not persuade us that the district court clearly erred in finding that the Bank fell short of proving that Vining-Sparks charged excessive commissions.

A securities broker's mark-up equals the price charged to the customer minus the prevailing market price. SEC Release, supra, at 16,631-26. A mark-up is excessive when it bears no reasonable relation to the prevailing market price. Ettinger v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 835 F.2d 1031 (3rd Cir.1987); Charles Hughes & Co. v. Securities & Exch. Comm'n, 139 F.2d 434, 437 (2nd Cir.1943), cert. denied, 321 U.S. 786, 64 S. Ct. 781, 88 L. Ed. 1077 (1944). To determine whether a mark-up on municipal securities bears a reasonable relation to the prevailing market price, the SEC turns to MSRB Rule G-30, see, e.g., In re Staten Securities Corp., Exchange Act Release No. 18628, 25 S.E.C. Docket 46 (April 9, 1982), which requires brokers to sell municipal securities at a price that is "fair and reasonable, taking into consideration all relevant factors, including the best judgment of the broker ... as to the fair market value ..., the expense involved in effecting the transaction, the fact that the broker ... is entitled to a profit, and the total dollar amount of the transaction." MSRB Manual--General Rules, G-30 (CCH) p 3646 (1985). Although both the SEC and the MSRB refuse to adopt a numerical standard, the SEC has consistently held that, in sales of equity securities, undisclosed mark-ups of more than 10 percent above the prevailing market price violate federal securities law. SEC Release, supra, at 16,631-25. Further, the SEC "also consistently has taken the position that mark-ups on debt securities, including municipal securities, generally are expected to be lower than mark-ups on equity securities, and has upheld NASD decisions finding mark-ups as low as 5.1% to violate the rules of the MSRB." Id. (footnotes omitted); see also Securities & Exch. Comm'n v. Charles A. Morris & Assocs., 386 F. Supp. 1327, 1334 n. 5 (W.D.Tenn.1973) ("it is the practice in the municipal bond industry to charge a retail customer a price which is no more than one-quarter of one per cent to five per cent over the then current market price for a bond") (cited with approval in In re Staten Securities Corp., Exchange Act Release No. 18628, 25 S.E.C. Docket 46 (April 9, 1982)).

Guided by MSRB Rule G-30 and SEC enforcement policies, particularly as detailed in the SEC Release, the district court found that the Bank had failed to prove that Vining-Sparks charged excessive commissions. The district court noted that only one mark-up exceeded 5 percent, a 5.02 percent mark-up on a group of MFRBs which Vining-Sparks had held in inventory for several days before selling to the Bank. In addition, the district court remarked that the NASD had thoroughly investigated the Bank's allegations of fraudulent mark-up practices, yet concluded that Vining-Sparks had not violated the MSRB's rules. Finally, the district court cited the testimony of Andrew Barnes, a former associate general counsel to the NASD. In Barnes's opinion, Vining-Sparks had not charged excessive mark-ups on the bonds sold to the Bank, primarily because Vining-Sparks's pricing practices were consistent with industry standards, yet also because of the difficulties encountered by a brokerage firm in buying and selling such unusual bonds.

After reviewing the evidence introduced by Vining-Sparks, the district court turned to the Bank's arguments, but found that the Bank had failed to cite any evidence, besides the mark-up percentage, that Vining-Sparks had charged the Bank an unfair price. Even though the Bank disputes this finding, it still points to no factor listed in MSRB Rule G-30, nor any other relevant factor, which supports its challenge to the district court's finding. Although an undisclosed mark-up of 5 percent on municipal bonds might sometimes violate Rule 10b-5, the Bank fails to convince us that the district court clearly erred when it found that a 5 percent mark-up on the bonds sold to the Bank, without more, was not so excessive as to require disclosure.

As a final matter, we address the Bank's contention that the district court clearly erred when it calculated some of the mark-ups charged by Vining-Sparks. Initially, we note that the district court properly held that the market price to which a broker adds a mark-up is the wholesale or inter-dealer price. See SEC Release, supra, at 16,631-26 to -27. For bonds purchased by Vining-Sparks through the J.J. Kenny Company, Inc. (J.J. Kenny), a municipal bond brokerage house, the district court simply subtracted the price paid by Vining-Sparks to acquire the bonds from the price it later charged the Bank, for the bonds were purchased from J.J. Kenny at wholesale. The Bank apparently concedes that the district court properly calculated Vining-Sparks's mark-ups on these bonds. For the remaining bonds, which Vining-Sparks purchased from its customers before reselling them to the Bank, the district court adjusted the price paid by Vining-Sparks to acquire the bonds to reflect a mark-down. The Bank contends that the district court clearly erred when it added back a mark-down to arrive at the prevailing market price.

We find ample support in the record for the district court's method of calculation. As the district court noted, the SEC Release instructs that the best evidence of a security's wholesale market price is the price one broker charges another. SEC Release, supra, at 16,631-26. Yet, where evidence of the inter-dealer price is not available, "the best evidence of the prevailing market generally will be the broker-dealer's contemporaneous retail purchases, adjusted to reflect the mark-down inherent in such customer transactions." Id. at 16,631-27 (footnote omitted). Further, Vining-Sparks's bond trader, Hickerson, testified that before offering to buy bonds from a customer, he would first determine the current wholesale price, then calculate an appropriate mark-down. If a customer agreed to sell at the resulting price, Vining-Sparks would buy the bonds. Finally, Barnes confirmed that Vining-Sparks's pricing practices, as described by Hickerson, conformed to industry standards. Thus, in light of the SEC Release and the testimony of both Hickerson and Barnes, the Bank fails to convince us that the district court made a mistake when it calculated Vining-Sparks's mark-ups.

The Bank next argues that the district court ignored its breach of contract claim. The Bank alleged that Vining had orally promised that the Bank would receive an accounting yield of 8 percent on its investments, and that the transaction confirmation statements showing yields to maturity of approximately 8 percent established a written contract, in which Vining-Sparks promised an 8 percent accounting yield. In addition, the Bank alleged that ViningSparks breached an implied covenant of good faith and fair dealing inherent in all contracts created under Kentucky law. In the district court's conclusions of law, however, the district court only addressed the claim for breach of an implied covenant of good faith and fair dealing, noting that the claim "merely restates the Bank's prior allegations of wrongdoing, about which no further discussion is necessary."

Although the district court did not expressly reject the Bank's claim that Vining and Vining-Sparks promised an accounting yield of 8 percent, the district court's ninety-eight pages of findings of fact and conclusions of law show beyond doubt that it implicitly rejected the claim. Where a district court has implicitly decided a narrow and specific issue, we will review the findings of fact and conclusions of law which necessarily support that decision, rather than remand for a certain express determination. Brown v. Baltimore & O.R.R., 805 F.2d 1133, 1141 (4th Cir.1986).

We find no error in the district court's implicit rejection of the Bank's breach of contract claim. Essentially, the Bank asserts that Vining and Vining-Sparks each guaranteed an 8 percent return on the Bank's investment. This assertion conflicts with the district court's findings. First, the district court found that the call features of each lot of bonds were disclosed, as were the risks those features presented. Second, the district court found that Vining-Sparks merely "expressed its opinion that [both the SFMRBs and the bonds with optional call features] would not be subjected to premature redemption during the holding period contemplated...."

The district court found that the Bank knew that the SFMRBs might be recalled if the underlying mortgages were prepaid. Indeed, the Bank tried to prove at trial that Vining had orally promised to lessen the risk of premature redemptions by only selling zero-coupon bonds which represented between 3 and 5 percent of each issue, and which the issuers would call last if the underlying mortgages were prepaid. Moreover, the Bank tried to show that Vining had misrepresented Vining-Sparks's research into prepayment speeds. Hence, the Bank's own allegations regarding the SFMRBs undermine its assertion that Vining and Vining-Sparks guaranteed a given yield, for they show, as the district court found, that the Bank knew that the rate of prepayment speeds would affect the bonds' yield.

Likewise, although the confirmation statements sent to the Bank following each sale of bonds with optional call provisions showed a yield to maturity of approximately 8 percent, they also showed that the bonds were callable, and furnished both the call date and call price. Each statement also contained a warning that "call features may exist which would affect yield." The district court found that the Bank understood the call features of these bonds, and the risks which the features posed, but bought the bonds anyway because of anticipated tax advantages. Thus, this finding similarly weakens the Bank's assertion that Vining and Vining-Sparks promised that the bonds with optional call provisions would produce a given yield.

While the Bank's knowledge that mortgage prepayment speeds might affect the yields of the SFMRBs, and that optional calls might affect the yields of the bonds with optional call features, casts serious doubt on its assertion that Vining and Vining-Sparks promised 8 percent yields, the district court's finding that Vining-Sparks had merely conveyed its "opinion" that the bonds would not be recalled prematurely shows beyond doubt that the district court rejected the Bank's breach of contract claim. To establish a claim for breach of a contract, a plaintiff must prove the existence of the contract. E.g., Johnson v. Davis, 516 S.W.2d 649 (Ky.1974). "A contract is a promise or set of promises for the breach of which the law gives a remedy, or the performance of which the law in some way recognizes as a duty." Restatement (Second) of Contracts § 1 (1979). Far from finding that Vining or Vining-Sparks promised that each lot of bonds would yield an 8 percent return on the Bank's investment, the district court found that they had, at most, predicted that the bonds would not quickly be redeemed. See Restatement (Second) of Contracts § 2 comment f (1979) ("A promise must be distinguished from a statement of opinion or a mere prediction of future events"). Implicit in this finding is the finding that Vining-Sparks did not promise an 8 percent yield. Accordingly, we agree with the district court's implicit rejection of the Bank's breach of contract claim.

Finally, the Bank asserts that the district court committed reversible error when it admitted into evidence the letter of caution which the NASD sent to Vining-Sparks following the NASD's investigation of the Bank's complaint, and a similar NASD letter of caution stemming from a complaint made by a Mississippi bank. The Bank objected to the letters' admission, contending that they were a "whitewash" by Vining's "good friends" on the NASD's District Business Conduct Committee, a committee on which Vining had formerly served. Noting that it understood the Bank's objection, the district court admitted the letters into evidence after remarking that the danger of giving the letters undue weight would not be so great in a bench trial.

The parties apparently agree that the letters contain hearsay statements. See Fed.R.Evid. 801. Yet, Vining-Sparks asserts that the letters are admissible under the public records and reports exception to the hearsay rule, Fed.R.Evid. 803(8)(C), whereas the Bank argues that the district court failed adequately to consider the trustworthiness of the letters before admitting them into evidence, as Rule 803(8)(C) demands. The Bank appears to concede that the NASD is a public agency as required to satisfy the exception.

Rule 803(8)(C) defines the public records and reports which the hearsay rule does not exclude as follows:

Records, reports, statements, or data compilations, in any form, of public offices or agencies, setting forth ... (C) in civil actions and proceedings ..., factual findings resulting from an investigation made pursuant to authority granted by law, unless the sources of information or other circumstances indicate lack of trustworthiness.

Opinions, conclusions, and evaluations, as well as facts, fall within the Rule 803(8)(C) exception. Beech Aircraft Corp. v. Rainey, 488 U.S. 153, 168-69, 109 S. Ct. 439, 449, 102 L. Ed. 2d 445 (1988). Further, the Advisory Committee notes that the rule "assumes admissibility in the first instance but with ample provision for escape if sufficient negative factors are present." In light of this presumption of admissibility, the party opposing the admission of the report must prove that the report is not trustworthy. Baker v. Elcona Homes Corp., 588 F.2d 551, 558 (6th Cir.1978), cert. denied, 441 U.S. 933, 99 S. Ct. 2054, 60 L. Ed. 2d 661 (1979).

We review the district court's decision to admit the letters into evidence for abuse of discretion. See Moss v. Ole South Real Estate, Inc., 933 F.2d 1300, 1305-06 (5th Cir.1991); Keith v. Volpe, 858 F.2d 467, 481-82 (9th Cir.1988), cert. denied, 493 U.S. 813, 110 S. Ct. 61, 107 L. Ed. 2d 28 (1989). Applying this standard, we hold that the district court did not err.

To determine whether a report is trustworthy, the Advisory Committee suggests four factors to consider: (1) the timeliness of the investigation, (2) the special skill or experience of the investigators, (3) whether the agency held a hearing, and (4) possible motivational problems. See Baker, 588 F.2d at 558 (applying the four listed factors in reviewing a district court's decision to admit a report).

Although the district court did not state fully its reasons for admitting the two letters, it surely considered and rejected the Bank's lengthy objection on the ground that the NASD had possible motivational problems springing from Vining's prior service on the District Business Conduct Committee. The district court, as noted in its findings of fact and conclusions of law, had already heard the testimony of Barnes, who had described in detail the process the NASD follows when investigating customer complaints. Thus, before ruling on the Bank's objection to the report, the district court had heard the testimony of a former associate general counsel to the NASD, Barnes, and a former member of the District Business Conduct Committee, Vining, and concluded that motivational problems did not render the report so untrustworthy as to require its exclusion.

The Bank also contends that because the NASD issued its letters of caution during the trial of this case, the first factor listed by the Advisory Committee favored excluding the letters. Yet, the first factor relates not so much to the time it takes to complete a thorough investigation, such as the one conducted by the NASD, but to the time between the acts or events which the agency is investigating and the start of the investigation. Baker, 588 F.2d at 558. The Bank cited no evidence showing that the NASD had failed to begin its investigation shortly after receiving the Bank's complaint.

As to the second factor, the district court failed before admitting the letters to say that the NASD's expertise and skill verified the letters' trustworthiness. In the district court's findings of fact, however, it noted that the NASD "has extensive expertise in analyzing the propriety of a member firm's mark-up practices." Therefore, it appears that the district court relied on its finding of expertise in deciding to admit the letters into evidence.

Finally, citing the third factor, the Bank asserts that the letters lack adequate trustworthiness because the NASD heard no sworn testimony or cross examination. Although the NASD did not conduct any hearings, we note, as we did in Baker, that the rule "makes no reference to such a requirement; the factor appears only to be one of those suggested by the Advisory Committee." 588 F.2d at 558.

In sum, we hold that the district court could reasonably have concluded that the letters did not lack trustworthiness. Thus, we refuse to find that the district court abused its discretion.

We find no merit to the Bank's grounds for appeal. Accordingly, we AFFIRM the judgment of the district court.