Walgreen Company (Walgreen) appeals a decision of the United States District Court for the District of Colorado. The district court denied Walgreen’s motion for remitti-tur, or in the alternative a new trial, on a jury award of $1.1 million in lost profits. The district court also denied Walgreen’s motion for judgment as a matter of law that Philip Oiness (Oiness) could not recover projected lost profits on the record evidence. Finally, the district court awarded Oiness *1028prejudgment interest at the Colorado statutory rate of 8%. Because the jury damages award was speculative, this court reverses-in-part, vaeates-in-part, and remands-in-part.
Oiness cross-appeals the trial court’s denial of its motion for enhanced damages and request for prejudgment interest at the rate of 16.52%. This court affirms the trial court on these issues.
BACKGROUND
This appeal features a small folding headrest called a Headchair. The Headchair resembles a tiny director’s chair which supports a user’s head off the ground. The Headchair sells primarily to sunbathers at the beach or to campers. This simple product has seen a tumultuous legal history.
Steven Younger (Younger) and Rudolf Fie-delak, co-inventors of the Headchair, filed a patent application for the Headchair on October 11, 1984. The patent issued on October 1, 1985 as U.S. Patent No. 4,544,203 (’203). Younger, Oiness, and other investors formed the corporation Sun Global Enterprises, Inc. (Sun Global). The inventors assigned the ’203 patent to Sun Global. In March 1984, Sun Global began making the Head-chair. Sales commenced in June 1984.
In mid-1985, B & E Sales Co. (B & E) began selling a folding headrest. Sun Global sued B & E in 1986 for infringement of the ’203 patent. In that trial, Younger testified that B & E’s infringement virtually destroyed the market for Sun Global Head-chairs. Sun Global won a lost profits damages award from B & E for harm to the retail market from 1986 through 1992.
In June 1986, Oiness formed a new corporation, Sun Products Group, Inc. (Sun Products), to sell the Headchair. Sun Products acquired the assets of the defunct Sun Global. Oiness thus acquired the ’203 patent. Oiness and Sun Products then sued Walgreen for infringement of the ’203 patent.
In April 1991, the district court held its first trial on this case. The jury found that Walgreen had willfully infringed the patent and awarded Oiness $300,000 in damages. Oiness v. Walgreen Co., 774 F.Supp. 1277, 21 USPQ2d 1654 (D.Colo.1991).
Walgreen appealed that decision to this court. Oiness v. Walgreen Co., 980 F.2d 742, 26 USPQ2d 1548 (Fed.Cir.1992) (non-precedential), cert. denied, 507 U.S. 1032, 113 S.Ct. 1849, 123 L.Ed.2d 473 (1993). On appeal, this court affirmed Walgreen’s liability, but vacated the damage award and remanded the case for reconsideration of the amount of damages. This court determined that the trial court had erred in its jury instructions on damages. The faulty instructions did not distinguish between the lost profits award for B & E’s infringement in 1985-1986 and the lost profits for Walgreen’s 1987-1990 infringement. Without this distinction, the jury could assess Walgreen for damages already compensated in the B & E trial. Furthermore, this court discerned a lack of substantial evidence to support the jury’s damage award. Id.
The district court retried the case on damages. In October 1994, the jury awarded Oiness $1,101,240 in lost profits and $10,150,-000 in projected lost profits. Specifically, the jury awarded projected lost profits of $1.4 million in the retail sales market, $750,000 in the premium market, and $8 million in the advertising specialty market.
Walgreen moved for judgment as a matter of law, remittitur of the damage award, and alternatively a new trial on damages, all of which the trial court denied in an Order Regarding Post-Trial Motions entered December 6, 1994. The district court also awarded 8% in prejudgment interest on the entire judgment.
DISCUSSION
Actual Sales
Walgreen moved for remittitur, or in the alternative a new trial and moved for judgment as a matter of law (JMOL) that the record evidence did not support projected lost profits. In considering a motion to amend the judgment, or in the alternative to grant a new trial on the amount of damages, a trial court must review the record to determine whether the jury’s verdict contravenes the “clear or great weight of the evidence.” Unisplay, S.A v. American Elec. Sign Co., *102969 F.3d 512, 517, 36 USPQ2d 1540, 1544 (Fed.Cir.1995) (quoting Standard Havens Prods., Inc. v. Gencor Indus., Inc. 953 F.2d 1360, 1367, 21 USPQ2d 1321, 1326 (Fed.Cir.1991), cert. denied, 506 U.S. 817, 113 S.Ct. 60, 121 L.Ed.2d 28 (1992)). This court reviews for an abuse of discretion the decision denying a grant of remittitur or a new trial because of an excessive damage award. K-B Trucking Co. v. Riss Int’l Corp., 763 F.2d 1148, 1162 (10th Cir.1985); see also Shearing v. Iolab Corp., 975 F.2d 1541, 24 USPQ2d 1133 (Fed.Cir.1992) (reviewing the denial of a motion for a new trial for an abuse of discretion).
In considering a motion for JMOL, a trial court reviews the record for substantial evidence to support the jury’s verdict. Federal Deposit Ins. Corp. v. United Pacific Ins. Co., 20 F.3d 1070, 1082 (10th Cir.1994).
On appeal from a decision on a motion for JMOL, this court must review the record for substantial evidence. Richardson v. Suzuki Motor Co., 868 F.2d 1226, 1240, 9 USPQ2d 1913, 1924 (Fed.Cir.), cert. denied, 493 U.S. 853, 110 S.Ct. 154, 107 L.Ed.2d 112 (1989). The measurement of actual damages for patent infringement is a question of fact. Brooktree Corp. v. Advanced Micro Devices, Inc., 977 F.2d 1555, 1578, 24 USPQ2d 1401, 1417 (Fed.Cir.1992).
Upon a finding of infringement, title 35 envisions the award of damages adequate to compensate the patentee for infringement, but in no event less than a reasonable royalty. 35 U.S.C, § 284 (1994). The patent owner bears the burden of proving this amount. SmithKline Diagnostics, Inc. v. Helena Lab. Corp., 926 F.2d 1161, 1164, 17 USPQ2d 1922, 1925 (Fed.Cir.1991). Beyond a reasonable or established royalty, a claimant must prove actual damages to establish entitlement to lost profits. Water Technologies Corp. v. Calco, Ltd., 850 F.2d 660, 671, 7 USPQ2d 1097, 1106 (Fed.Cir.), cert. denied, 488 U.S. 968, 109 S.Ct. 498, 102 L.Ed.2d 534 (1988). Proof of actual damages must include a causal connection between the infringement and the lost profits. Rite-Hite Corp. v. Kelley Co., 56 F.3d 1538, 1585, 35 USPQ2d 1065, 1102 (Fed.Cir.) (en banc), cert. denied, - U.S. -, 116 S.Ct. 184, 133 L.Ed.2d 122 (1995). “To recover lost profits as actual damages, a patent holder must demonstrate that there was a reasonable probability that, but for the infringement, it would have made the infringer’s sales.” Minnesota Mining & Mfg. Co. v. Johnson & Johnson Orthopaedics, Inc., 976 F.2d 1559, 1577, 24 USPQ2d 1321, 1336 (Fed.Cir.1992) (citing State Indus., Inc. v. Mor-Flo Indus., Inc., 883 F.2d 1573, 1577, 12 USPQ2d 1026, 1028 (Fed.Cir.1989), cert. denied, 493 U.S. 1022, 110 S.Ct. 725, 107 L.Ed.2d 744 (1990)).
The district court properly instructed the jury that Oiness was entitled to present proof of lost profits based upon Walgreen’s sales. Oiness, however, did not present proof of Walgreen’s sales. Instead Oiness presented only pictures of Headrest displays in three Walgreen stores. From these pictures, Oiness presumed that Walgreen devoted one and a half to two square feet of floor space in every store to Headrest sales. Without proof that every store sold Headrests or the amount of space in those stores for Headrest sales, Oiness then multiplied its estimated square footage by 1600, the number of Walgreen stores. Oiness next relied on a Value Line report for investors which discussed the vague concept of sales per square foot. This report suggested that Walgreen averaged sales of $379 per square foot of floor space. This report made no assessment of the sales of Headrests and no assessment of whether Headrest sales fell above or below the square foot sales average. Nonetheless, Oiness multiplied the presumed floor space in each Walgreen store by $379 per square foot. Finally, Oiness projected this calculation over a five year period. Thus Oiness extrapolated that Walgreen would have sold 3.6 to 4.8 million Headrests. This number of sales generates lost profits between $4,840,000 and $6,450,000.
This evidence adds vague estimation and gross extrapolation to unsupported presumption. At every step, this damages calculation is fraught with speculation. Three pictures do not support a conclusion for 1600 stores. The Value Line report does not supply any *1030information about Headrest sales. The five-year projection does not account for market fluctuations over time and does not relate to the three pictures or the Value Line report at all. Instead of presenting evidence of actual sales combined with reliable economic analysis of demand, supply, and price over time, Oiness invites the jury to engage in rapt speculation. Based as it is on insubstantial evidence, this court reverses the jury’s $1,101,240 lost profits award.
A court is not at liberty to supplant its own judgment on the damages amount for the jury’s findings. Unisplay, 69 F.3d at 519. Therefore, in holding that a jury damage award is excessive, an appellate court has two options. It may simply reverse the jury award and order a new trial or allow plaintiff the option of agreeing to a remittitur in a specified amount. See id.; 11 Charles A. Wright, Arthur R. Miller, Mary Kay Kane, Federal Practice & Procedure § 2820 (2d ed. 1995). This court has adopted the “maximum recovery rule” which requires this court to remit the damage award to the highest amount the jury could “properly have awarded based on the relevant evidence.” Unisplay, 69 F.3d at 519.
The record shows conclusively that Walgreen sold 142,230 Headrests in the retail market. Younger testified that the average net profit on the Headchair in the retail market was $1.33. Multiplying the amount of Headrests Walgreen sold by the admitted average net profit of the Headchair equals a total of $189,000 in the retail market.
In addition, Vernon Brunner (Brunner), the Executive VP of Marketing for Walgreen, testified that at the end of August 1990, Walgreen had 23,610 Headrests left in its stores. The warehouse report, which was the cornerstone of Walgreen’s proof of Headrest sales, shows no movement of Headrests after this date. The 142,230 units reflect the 165,840 units moved to stores, minus the inventory remaining in the stores as of August 1990. Walgreen produced no records accounting for the whereabouts of the 23,610 Headrests. Brunner also testified that the stores were not advised to stop selling the Headrest until approximately March 1991. It is uncertain what happened to the 23,610 Headrests. Therefore, in view of Walgreen’s lack of records, it was reasonable that the jury inferred these Headrests had been sold. Lam, Inc. v. Johns-Manville Corp., 718 F.2d 1056, 219 USPQ 670 (Fed.Cir.1983). Adding the 23,610 unaccounted for Headrests with the 142,230 sold amounts to $220,567 in actual sales. This amount serves as a benchmark for remittitur.
Oiness questioned this evidence on actual sales by alleging that Walgreen withheld evidence of actual sales. To the contrary, Walgreen presented inventory and warehouse records. Brunner, presented un-rebutted testimony that these records were the most accurate method to determine the number of Headrests Walgreen had available for sale. Moreover, in light of the utter failure of Oiness to present proof of lost profits, the record contains no better evidence of Walgreen’s actual sales. Because this evidence is the only reliable record of actual sales, this court vacates and remands to the trial court with instructions to order judgment in the amount of $220,567. Should Oiness refuse to accept the reduction of its award, the trial court may grant Oiness a new trial on damages.
This result is consistent with Lam. In Lam, this court inferred causation of damages when the patent owner and the infringer were the only suppliers of the product. With respect to reliance on the infringer’s sales records, the court noted that “any adverse consequences must rest on the infringer when the inability to ascertain lost profits is due to the infringer’s own failure to keep accurate or complete records.” Lam, 718 F.2d at 1065 (citing Milgo Elec. Corp. v. United Business Communications, Inc., 623 F.2d 645, 665, 206 USPQ 481, 496 (10th Cir.), cert. denied, 449 U.S. 1066, 101 S.Ct. 794, 66 L.Ed.2d 611 (1980)). The Lam court, however, noted that the plaintiff has the burden of showing with reasonable probability defendant’s sales. 718 F.2d at 1065. Even where defendant’s records are not complete, damages may “not be determined by mere speculation or guess.” Id. (quoting Story Parchment Co. v. Paterson Parchment Paper Co., *1031282 U.S. 555, 563, 51 S.Ct. 248, 250, 75 L.Ed. 544 (1931)).
In this case, Oiness purported to meet its burden of proof with mere speculation and guess work. Any insufficiency in Walgreen's records cannot supplant Oiness's burden to prove lost profits by a preponderance of evidence. See SmithKline, 926 F.2d at 1164 (at trial it is the patentee's burden to prove lost profits by a "preponderance of the evidence"). Due to Oiness' failure of proof, the best remaining evidence is Walgreen's records. Thus, this court remands with the above instructions.
Projected Lost Profits
This court upholds a jury's damages award unless "grossly excessive or monstrous," clearly not supported by evidence, or based only on speculation or guesswork. Brooktree, 977 F.2d at 1580 (citing Los Angeles Memorial Coliseum Comm'n v. NFL, 791 F.2d 1356, 1360 (9th Cir.1986), cert. denied, 484 U.S. 826, 108 S.Ct. 92, 98 L.Ed.2d 53 (1987)). In this regard, this court acknowledges that a patentee may produce sufficient evidence to recover projected future losses, see Lam, 718 F.2d at 1068, but those projections must not be speculative. See Sunward Corp. v. Dun & Bradstreet, Inc., 811 F.2d 511, 541 (10th Cir.1987); Brooktree, 977 F.2d at 1581 (citing Bio-Rad Lab., Inc. v. Nicolet Instrument Corp., 739 F.2d 604, 616, 222 USPQ 654, 664 (Fed.Cir.), cert. denied, 469 U.S. 1038, 105 S.Ct. 516, 83 L.Ed.2d 405 (1984)). "The burden of proving future injury is commensurately greater than that for damages already incurred, for the future always harbors unknowns." Brooktree, 977 F.2d at 1581. `While estimates of lost future profits may necessarily contain some speculative elements," Malloy v. Monahan, 73 F.3d 1012, 1016 (10th Cir.1996) (citing United Steelworkers of America v. CCI Corp., 395 F.2d 529, 533 (10th Cir.1968), cert. denied, 393 U.S. 1019, 89 S.Ct. 627, 21 L.Ed.2d 564 (1969); United States v. Griffith, Gornall & Carman, Inc., 210 F.2d 11, 13 (10th Cir.1954)), "the factflnder must have before it `such facts and circumstances to enable it to make an estimate of damage based upon judgment, not guesswork.'" Id. (quoting Griffith, Gornall & Carman Inc., 210 F.2d at 13).
This case involves three different markets: the premium market, the retail market, and the advertising specialty market. Oiness contends that Waigreen's infringement destroyed all three markets. The jury awarded Oiness projected lost profits in all three of these markets.
A.
The retail market concerns products sold at retail through department, stores, drug stores, chains, and direct sales to consumers. Douglas MacLachian (IviacLachian) was Oiness' expert in the retail market. MacLachian developed his forecast in this market based on information from Sun Product's attorneys, accountant, and officers. MacLachlan forecasted Sun Product's lost profits from the start of infringement until 2002, the patent expiration date.
MacLachian admitted that he found very little sales history to support his projections. The prior infringement by B & E complicated his reliance on the sales history of the Headchair. Due to this factor, MacLachlan relied on the sales unit figures for the second half of 1984, 30,746, and the first half of 1985, 304,045-the only figures before B & E's infringement. He determined that the difference between these figures represented a 900% increase. He used this initial burst in product sales to project the growth rate for sales over the duration of the patent absent infringement. Recognizing the alarming size of this growth rate, however, he proposed "the conservative approach" of reducing this growth rate to a "mere" 450% increase in future sales for several years, with another reduction to 150% growth near the end of the patent term. In other words, MacLachian still projected more than a doubling of annual sales near the end of Oiness' patent term.
MacLachian's entire premise is flawed. MacLachlan relies on figures from the first months after Headchair's introduction into the market place. He then assumes that those figures, from a few months, support an extrapolation of demand over future years. The economic axiom that demand curves *1032slope downward calls into question the doubling of demand for a product year after year. MacLachlan offered no sound economic reasoning to support his assumption that Headchair sales would quadruple and double throughout the life of the patent. The figures for the first burst of sales do not supply adequate support for his assumption.
MacLachlan’s assumption is also flawed because he had no evidence to show that the 1984-1985 sales actually represented sales to customers over this brief period, as opposed to large initial sales to retail outlets which might have languished on store shelves for months. In sum, MacLachlan’s projections lacked evidentiary support.
In Lam, this court allowed projected lost profits in a two-supplier market upon a showing of actual pre-infringement and post-infringement growth rates. Lam, 718 F.2d at 1068. Oiness neither established a reliable pre-infringement nor a reliable post-infringement growth rate. The 1984-1985 start-up figures distort MacLachlan’s growth rate projections. Furthermore, MacLachlan admits he did not conduct any market surveys of the Headchair. Rather, he relied on his personal conjecture that Sun Products would employ manufacturing representatives around the country at trade shows to increase distribution over the life of the patent.
Because MacLaehlan’s projections of lost profits in the retail rest on faulty assumptions and a lack of reliable economic testimony relevant to this market, this court determines that the record does not support the jury’s award of $1.4 million. The only record testimony was bald speculation by Oiness’ expert. Therefore, this court reverses the jury’s award of projected lost profits in the retail market.
B.
The advertising specialty market concerns products — often calendars and pens — with a company’s logo which are usually given away to promote the company’s business reputation. The advertising specialty market operates by engaging a manufacturing representative to sell the product to a distributor. The distributor then resells the product to an interested company. The distributor must obtain the product for a price lower than the interested company might obtain it elsewhere. Therefore, Oiness argues, when Walgreen sold the Headrest for $1.00, the price Sun Products could demand in the advertising specialty market plummeted.
Oiness’ expert in the advertising specialty market was Glen Holt (Holt). Holt admitted that he did not take into account B & E’s infringement when projecting lost profits in this unique market. This oversight is fatal to Holt’s predictions. To the extent B & E had already damaged pricing in this market, Walgreen cannot be liable for damage caused by another infringer.
In the B & E infringement trial, Younger testified that B & E virtually destroyed the market for the Headchair. In his 1988 testimony, Younger held B & E responsible for a 90% drop in the business of Sun Global. B & E’s infringement extended into 1986 — less than a year before Walgreen’s infringement. Because Holt did not consider the effect of B & E’s infringement on the Headchair market, his projections of Walgreen’s harm to the price of Headehairs, less than a year later, are suspect.
Just as significantly, Holt did no market analysis to determine if market forces, other than infringement, might have forced the price of Headehairs to dip. For instance, Oiness projected vast market penetration in other markets in the same period. If this had occurred, the market itself might have reduced the price as much or more than Holt attributed to Walgreen’s infringement. Without more reliable market analysis, Holt provided no more than speculation upon which the jury cannot base an award. Because Oiness did not show an established business in the advertising specialty market or a reliable basis for Holt’s predictions, this court reverses the jury award on projected lost profits in the advertising specialty market.
C.
In the premium market, customers buy the products to give away or sell as *1033premiums in connection with another product or service, such as a prize in a cereal box. Customers may also use premium products as incentives for dealers or sales persons to sell more merchandise.
George Kling (Kling) was Oiness’ expert witness in the premium market. He testified that this market relies heavily on the perception of the value of the premium product. He noted that the perceived value of the Headchair dropped because of Walgreen’s infringement. Before Walgreen’s activity the perceived value of the Headchair was $5.00 or $6.00; it later dropped to $1.00. Because of the drop in the perceived value of the Headchair, Kling testified, the premium representatives would no longer carry the Headchair as a premium product.
Once again, Kling did not factor the B & E infringement into his calculation of the damage to the perceived value of the Headchair. Without this calculation, Fling's calculations run the risk of charging Walgreen with damage caused by another’s infringement.
The premium market was a new market for Sun Products. Sun Global had not entered this market. Nonetheless, the value of the Headchair in one market is not independent of its value in other markets. Without consideration of the B & E infringement on Headchair value, Oiness has not shown that the premium market remained unaffected by B & E’s infringement.
Edgar Taylor (Taylor), director of the Nestle Food Company Baking Division at Carnation, testified that Carnation used the Headchair in its promotion of the “Carnation Do it Yourself Diet Plan.” He testified that the reason Carnation chose the Headchair as a promotional product was its price. Taylor also testified that he would stay away from a product if “he thought the perceived value would decline during the course of the program.” Kling testified that, after Walgreen’s infringement, Carnation withdrew the Head-chair as one of its promotional products.
The mere temporal association of Carnation’s action and Walgreen’s infringement does not suffice to show causation. No record evidence showed that Carnation terminated its premium product due to price decreases caused by Walgreen. The record does not show a link between Walgreen’s infringement and price decreases in the He-adchair. B & E’s infringement or other market forces may also have caused the decreases. Without credible economic testimony, this court cannot permit a jury to base its award on speculation. This court reverses this projection as well.
Interest
The trial court awarded Oiness 8% in prejudgment interest on the entire damages award. This court reviews grant or denial of prejudgment interest for an abuse of discretion. Lummus Indus., Inc. v. D.M. & E. Corp., 862 F.2d 267, 274, 8 USPQ2d 1983, 1988 (Fed.Cir.1988). Under 35 U.S.C. § 284, prejudgment interest ensures adequate compensation for the infringement. General Motors Corp. v. Devex Corp., 461 U.S. 648, 655-56, 103 S.Ct. 2058, 2062-63, 76 L.Ed.2d 211, 217 USPQ 1185, 1188-89 (1983); Underwater Devices, Inc. v. Morrison-Knudsen Co., 717 F.2d 1380, 1389, 219 USPQ 569, 576 (Fed.Cir.1983). Prejudgment interest has no punitive, but only compensatory, purposes. Interest compensates the patent owner for the use of its money between the date of injury and the date of judgment. Bio-Rad Lab., Inc. v. Nicolet Instrument Corp., 807 F.2d 964, 969, 1 USPQ2d 1191, 1195 (Fed.Cir.1986), cert. denied, 482 U.S. 915, 107 S.Ct. 3187, 96 L.Ed.2d 675 (1987).
The trial court abused its discretion by awarding prejudgment interest on Oiness’ entire damages award, including the projection of future damages. By awarding interest on projected lost profits, the trial court compensated Oiness for losses it had not yet suffered. In other words, the court granted Oiness interest for the use of its money when Oiness’ money had not been used. This award violates the compensatory purpose of prejudgment interest. Therefore, this court reverses the trial court’s grant of prejudgment interest on the future damages portion of Oiness’ damage award. Upon remand, the trial court may award interest on the $220,-567 award of lost profits supported by the record evidence.
*1034 CONCLUSION
This court remands on the issue of actual damages because the jury award exceeded Walgreen’s sales in the retail market. This court reverses the jury award for projected lost profits in the retail, advertising specialty, and premium markets. The jury based these awards on speculation. The record does not support these awards. This court also reverses on the trial court’s granting of prejudgment interest on any damages based on projections.
COSTS
Each party shall bear its own costs. AFFIRMS-IN-PART, REVERSES-IN-PAJRT, REMANDS-IN-PART, and VACATES-IN-PART.