George Lahood pleaded guilty to one count of conspiracy to commit bank fraud in violation of 18 U.S.C. § 371. He attacks the amount of loss used to calculate his sentence. Finding no error, we affirm.

I.

Lahood and his wife own Eaglevision Communications Products, Inc. (“Eaglevision”). Lahood had signatory authority on the corporation’s checking account at Red River Bank. Iyad Samara operated several businesses with checking accounts at Cottonport Bank: and Union Bank and had signatory authority on the accounts. Samara enlisted Lahood and two others in a check kiting scheme, 1 and from June 2002 through January 2003 Lahood and Samara regularly exchanged checks and money orders to inflate their account balances. During this time, Lahood conducted transactions totaling approximately $1,000,000.

The check kiting scheme was discovered in January 2003, and in February 2003 *995 Red River Bank filed a Suspicious Activity-Report regarding the Eaglevision account. At the time of the discovery Samara’s account at Cottonport Bank was overdrawn by $52,000 as a result of the scheme; La-hood’s account had no overdraft.

A portion of the $52,000 overdraft was collected from Samara’s other accounts at Cottonwood Bank; he paid the remainder with $10,000 cash and a $23,000 loan. At the time of sentencing the balance on the loan was approximately $6,200.

II.

Lahood was named in three counts of a 118-count indictment: one count of conspiracy to commit bank fraud in violation of 18 U.S.C. § 371 and two counts of bank fraud in violation of 18 U.S.C. §§ 1344 and 2. He pleaded guilty to the conspiracy charge.

The presentence investigation report (“PSR”) identified a $52,000 loss from the conspiracy. Lahood objected to the loss amount, and the district court received oral argument and witnesses at sentencing. Testimony was heard from FBI agent Randolph Deaton and Beth Pucheu, the internal auditor of Cottonwood Bank.

The district court found that the amount of loss for sentencing purposes was at least $33,000 because the actual loss by the bank was covered by a $10,000 cash deposit and a $23,000 loan. 2 This resulted in an offense level of 12, which was adjusted to 10 for acceptance of responsibility. The guideline range was 6 to 12 months’ imprisonment, and Lahood was sentenced to 8 months’ incarceration followed by 3 years’ supervised release, plus a $10,000 fine and restitution of $6,200.

III.

Lahood makes three arguments regarding the loss amount. First, not all the loss was attributable to his actions. Second, the court selected the wrong date on which to calculate the amount. Finally, even if the correct date was used, based on the testimony at sentencing the amount should be $23,000.

We review de novo the application of the sentencing guidelines, but we review factual findings for clear error. United States v. Haas, 171 F.3d 259, 268 (5th Cir. 1999). A finding of fact is not clearly erroneous “[a]s long as it is plausible in light of the record read as a whole.” United States v. Betancourt, 422 F.3d 240, 245 (5th Cir. 2005) (quoting United States v. Morris, 46 F.3d 410, 419 (5th Cir. 1995)).

In calculating the loss caused by fraud, the sentencing court “need only make a reasonable estimate.” U.S.S.G. § 2B1.1 comment. (n.3(C)). 3 We give the district court wide latitude to determine the amount of loss, United States v. Cothran, 302 F.3d 279, 287 (5th Cir. 2002), because “the sentencing judge is in a unique position to assess the evidence and estimate the loss based upon that evidence.” U.S.S.G. § 2B1.1 comment. (n.3(C)). The determination of the loss amount is a factual finding and thus is shielded by the clearly erroneous rule on appeal. United *996 States v. Glinsey, 209 F.3d 386, 393 (5th Cir. 2000). 4

A.

Lahood contends that of the loss amount in the PSR, only $16,000 is directly attributable to his actions. We disagree.

Lahood pleaded guilty of conspiracy to commit bank fraud. Under the sentencing guidelines the conduct of others can be used to determine the sentencing range if such conduct is reasonably foreseeable and in furtherance of jointly undertaken criminal activity. U.S.S.G. § lB1.3(a)(l)(B). The goal of the conspiracy was to defraud banks with a check kiting scheme. The scheme succeeded, and check kiting by conspiracy members resulted in a loss to Cottonport Bank. Even if the particular loss was entirely on account of the actions of Samara or other conspiracy members, it can properly be used to sentence Lahood because the conduct — check kiting — was both reasonably foreseeable and in furtherance of the conspiracy.

B.

Lahood argues that the loss amount must be calculated on the date he withdrew from the conspiracy, which he claims is January 30, 2003. This argument is unavailing.

The amount of loss resulting from a check kiting scheme is measured at the time the scheme is discovered. United States v. Frydenlund, 990 F.2d 822, 825-26 (5th Cir. 1993). Restitution following the scheme’s discovery does not warrant a decrease in the loss amount. Id. at 826.

Deaton, who investigated the case for the FBI, testified that at the time of the scheme’s discovery the overdraft on Sara-ma’s account was approximately $52,000, which was partially recovered from other accounts. The remaining overage was paid with a $10,000 cash deposit and a loan for “20,000 and change.” Pucheu testified that the check kite was discovered at the end of January 2003 and that on January 28 Samara’s account had a shortfall of $33,000 based on the check-kiting scheme. The shortfall was repaid with a $10,000 cash deposit and $23,000 loan. Lahood did not present testimony contradicting these facts, but on cross-examination he elicited responses showing that the amount of overdraft on January 30 was never measured.

The correct time to measure the loss is not the date on which Lahood withdrew from the conspiracy, but the date when the kite is discovered. Also, we have never held that the loss must be calculated as of the exact date of the discovery. 5 Viewing the record as a whole, one will conclude that it is plausible that a reasonable estimate of the loss at the time of the scheme’s discovery was $33,000, so there is no clear error.

C.

Lahood argues that even if the district court calculated the loss as of the proper date, the loss amount should be $23,000, because Pucheu testified that “it is fair to say that the bank was only out $23,000.” This attempt to alter the meaning of the testimony by removing it from context fails: Pucheu maintained throughout her *997 testimony that the loss amount was $33,000, which was recovered after the scheme’s discovery by Samara’s $10,000 cash deposit and $23,000 loan. 6

D.

The government asks us to hold that the loss amount for sentencing purposes is the float at its highest point during the course of the check kite. 7 Because it is not squarely before us in this case, we need not, and do not, decide that question.

The judgment of sentence is AFFIRMED. 8