REVISED October 24, 2016

        IN THE UNITED STATES COURT OF APPEALS
                 FOR THE FIFTH CIRCUIT
                                                               United States Court of Appeals
                                                                        Fifth Circuit
                                   No. 15-10805                       FILED
                                                               October 05, 2016

UNITED STATES OF AMERICA,                                        Lyle W. Cayce
                                                                      Clerk
            Plaintiff - Appellee

v.

VINCENT BAZEMORE,

            Defendant - Appellant




                Appeal from the United States District Court
                     for the Northern District of Texas


Before KING, SMITH, and COSTA, Circuit Judges.
PER CURIAM:
      Defendant–Appellant Vincent Bazemore was convicted of mail fraud for
his part in a scheme to procure life insurance policies by misrepresenting the
applicants’ net worths and their intention to transfer the policies to a third
party. This court previously affirmed Bazemore’s conviction but vacated his
sentence and the restitution order. On remand for resentencing, the district
court applied an 18-level enhancement to Bazemore’s base offense level due to
the actual loss caused by Bazemore’s scheme to insurers and a lender.
Bazemore again appeals his sentence, raising several challenges to the district
                                       No. 15-10805

court’s application and calculation of the actual loss enhancement. For the
following reasons, we AFFIRM the district court’s sentence in full.
             I.     FACTUAL AND PROCEDURAL BACKGROUND
       This appeal arises out of an insurance fraud scheme perpetrated by
Defendant–Appellant Vincent Bazemore.                In resolving Bazemore’s original
appeal, we described the scheme and procedural history in detail, United
States v. Bazemore (Bazemore I), 608 F. App’x 207, 209 (5th Cir. 2015), and we
now recount the scheme and procedural history as relevant to the sentencing
question before us today. “Bazemore’s scheme involved tricking insurance
companies into issuing stranger-owned (or originated) life insurance (“STOLI”)
policies to unqualified applicants.” 1         

convinced senior
citizens of relatively modest means to apply for multi-million dollar life
insurance policies intended for high net-worth individuals. 

the policies by grossly inflating the applicants’ net worths on the policy
applications and falsely claiming that the applicants did not intend to transfer
the policy to a third party 2 and that the premiums would not be financed by a
third party. 

did not misrepresent the applicants’ age
or health status on any application. 

the first two years of
the policy premiums using loan proceeds from a lender, Portigon AG, 3 at which
point he planned to sell the policy to a third party investor, use the proceeds to



       1  STOLI policies are life insurance policies held by a third party who has no insurable
interest in the insured. Bazemore I, 608 F. App’x at 209. As discussed in Bazemore’s
presentence report (“PSR”), “STOLI policies are not illegal; however they circumvent state
insurable interest laws and are inconsistent with the established and legitimate purposes of
life insurance.”
        2 “The insurers that issued policies to Bazemore’s applicants would, without exception,

deny life insurance policies to applicants that intended from the outset to transfer the policy
to a third party” who had no insurable interest in the insured. 

the loan proceeds originated with Portigon, the First Bank of Delaware

actually issued the loans to Bazemore because Portigon did not have a retail license to make
direct loans.

                                              2
                                  No. 15-10805

repay the loan, and share the remainder with the applicant. 

policy issued, Bazemore, in his role as an insurance agent, received a
commission roughly equivalent to the cost of the first year’s premium. 

result of this scheme, “Bazemore was charged and convicted of four
counts of mail fraud, each relating to a STOLI policy for which he received a
commission payment.” 

At Bazemore’s first sentencing, “[t]he
district court calculated a [G]uidelines range of 292 to 365 months’
imprisonment based on an offense level of 39 and criminal history category of
II.”   

“The offense level was largely the product of a 24-point
enhancement for the scheme’s intended loss to the insurers, which the district
calculated to be $81 million, the sum of the death benefits for all of the policies
issued to Bazemore’s applicants.” 

court also calculated that
Bazemore owed restitution of $4,014,627.13. 

was the sum of
two distinct amounts: (1) an actual loss of $2,266,665.13 suffered by insurers
who paid commissions to Bazemore, and (2) an actual loss of $1,747,962
suffered by Portigon. 

these findings, the district court sentenced
Bazemore to 292 months’ imprisonment and ordered restitution of
$4,104,627.13. 

his conviction, sentence, and restitution order. 

affirmed Bazemore’s conviction but vacated his sentence and the
restitution order. 

As to his sentence, the court found that the
district court erred in using the sum face value of the insurance policies—$81
million—to calculate the intended loss from Bazemore’s scheme. 

14. The court noted that Bazemore made no misrepresentation as to the
applicants’ age or health status and this mitigated some of the potential harm
from the fraud. 

Accordingly, it concluded that the district court
could not apply an intended loss enhancement based on the $81 million sum


                                        3
                                       No. 15-10805

face value unless the Government proved by a preponderance of the evidence
that the fraudulent policies posed a risk of financial loss to the insurers that
the same policies issued to qualified insureds did not. 

As to the
restitution order, the court found that the district court incorrectly calculated
the actual loss amounts suffered by the insurers.                

    The court
instructed that the formula to use for actual loss on a rescinded STOLI policy
(in the restitution context) on remand was “the commission the insurer paid to
Bazemore less any premium payments that it retained.” 

the probation officer issued an addendum to the presentence
report (“PSR”), concluding that actual loss, rather than intended loss, should
be used to calculate the loss enhancement at resentencing. For purposes of the
loss enhancement, the PSR calculated a total actual loss of $1,282,636 to the
insurers targeted by Bazemore, using the formula described by the court in
Bazemore I (commissions paid by the insurers less any premium payments
they retained). 

The PSR also calculated an actual loss of
$1,747,962 to the lender, i.e., the same loss calculated by the district court for
Portigon at Bazemore’s first sentencing. But the PSR noted that Portigon
transferred its loan and securities portfolio to EAA PF LLP in January 2015,
so EAA replaced Portigon as the victim of Bazemore’s scheme. 4


       4 Through a complicated series of transactions, the details of which are not relevant
to the issues in this appeal, EAA has assumed what appears to be an identical role to that
previously held by Portigon and now possesses the interest in the loans originally used to
finance Bazemore’s insurance fraud scheme. Specifically, Portigon during the first
sentencing (and EAA during the resentencing) held the interest in thirteen insurance
premium loans. At the time of the first sentencing, nine of the loans were inactive either
because (1) the insurer had rescinded the underlying life insurance policy and refunded the
premium, resulting in Portigon recovering the loan proceeds, interest, and fees; (2) Portigon
had reached agreements with the policyholders for repayment of the loans, resulting in
Portigon recovering most, but not all, of the loan proceeds; or (3) the insurance policies had
lapsed for failure to pay the premiums, resulting in Portigon’s failing to recover any of the
loan proceeds. Four of the loans, however, remained active and were still being used to pay
the premiums on their related life insurance policies.

                                              4
                                      No. 15-10805

       On August 17, 2015, the district court held Bazemore’s resentencing
hearing. The district court adopted the PSR’s calculation of actual loss to the
insurers, finding they suffered a total actual loss of $937,612. 5 The district
court then calculated the actual loss to Bazemore’s lender. Looking to Fifth
Circuit caselaw, the court found that actual loss should be calculated “at the
time of the original sentencing,” thereby ignoring Portigon’s transfer of the
loans (and its related security interest in the insurance policies) to EAA
because the transfer occurred after the first sentencing. The district court
included only the nine loans that had been rescinded, settled, or lapsed in
calculating the lender’s total actual loss of $1,747,962. The district court did
not include any actual loss for the four loans related to active policies, noting
that the Government stated that any loss for those loans was “speculative.”
       Because the loss to the insurers and the lender combined for a total
actual loss of $2,685,574, the district court applied an 18-level enhancement.
See U.S. Sentencing Guidelines Manual § 2B1.1(b) (U.S. Sentencing Comm’n
2012) (providing for an 18-level enhancement for offense causing losses
between $2.5 and 7 million). The district court calculated a Guidelines range
of 151 to 188 months’ imprisonment based on a total offense level of 33 and a
criminal history category of II.         Similar to the first sentence, Bazemore’s
offense level was largely the product of the loss enhancement. The court
sentenced Bazemore to 188 months’ imprisonment, ordered restitution in the
amount of $2,685,574, and further explained that it would have imposed the
same sentence even if it had not overruled Bazemore’s objections. Bazemore
timely appeals the actual loss enhancement to his base offense level.




       5 The district court excluded the actual loss suffered by one of the insurers because
the evidence of that loss was incomplete at the time of the first sentencing.

                                             5
                                 No. 15-10805

                 II.   THE ACTUAL LOSS CALCULATION
A. The law of the case doctrine and the mandate rule
      On appeal, Bazemore argues, as he did before the district court, that the
district court erred in applying a loss enhancement based on a finding of actual
loss because both the law of the case doctrine and the mandate rule compelled
a finding of zero actual loss. Under the law of the case doctrine, “an issue of
fact or law decided on appeal may not be reexamined either by the district court
on remand or by the appellate court on a subsequent appeal.” United States v.
Matthews, 

, 657 (5th Cir. 2002) (quoting Tollett v. City of Kemah,

, 363 (5th Cir. 2002)). “[A] corollary or specific application of the
law of the case doctrine” is the mandate rule. United States v. Pineiro, 

, 205 (5th Cir. 2006) (per curiam). That rule “prohibits a district court
on remand from reexamining an issue of law or fact previously decided on
appeal and not resubmitted to the trial court on remand.” 

[mandate] rule bars litigation of issues decided by the district court but
foregone on appeal or otherwise waived, for example because they were not
raised in the district court.” United States v. Lee, 

, 321 (5th Cir.
2004). Accordingly, a district court cannot “reconsider its own rulings made
before appeal and not raised on appeal.” 18B Charles Alan Wright et al.,
Federal Practice and Procedure § 4478.3 (2d ed. 2016). Both the law of the case
doctrine and the mandate rule apply to issues decided expressly or implicitly.
See 

; 

. “We review de novo a district
court’s application of the remand order, including whether the law-of-the-case
doctrine or mandate rule forecloses the district court’s actions on remand.”
United States v. Carales–Villalta, 

, 344 (5th Cir. 2010); accord

.




                                       6
                                  No. 15-10805

      Bazemore’s argument rests on two objections that he made at his first
sentencing. In these objections Bazemore argued that (1) actual loss to the
insurers for the commissions they paid Bazemore should be calculated based
on losses to the insurers rather than gains to Bazemore because gains should
only be used as an alternative measure of loss when loss could not be
determined (Objection No. 2); and (2) any actual loss suffered by the insurers
in entirely refunding the premiums was not “reasonably foreseeable” to
Bazemore and thus “there was no actual loss in this case” (Objection No. 3). In
addressing the two objections the district court stated:
             With respect to Mr. Bazemore’s Objection No. 2 that
      commission payments were not losses to the insurance
      company, . . . .
             [Guidelines] Application Note 3(B) defines gain and [says
      that gain] applies as an alternative measure of loss only if there is
      a loss but it reasonably cannot be determined. Mr. Bazemore says
      you should not be able to use gain [as a measure of the loss he
      caused] . . . under Application Note 3(B) when you can ascertain
      the loss . . . .
             I will sustain Mr. Bazemore’s objection to this figure [i.e., the
      gains to Bazemore from commissions]. But I will note that it does
      not change the loss amount. That means I will not include this
      figure in the loss amount under Mr. Bazemore’s Objection No. 3.
Bazemore argues that the district court’s rulings on these objections show that
it found zero actual loss for purposes of determining a loss enhancement, and
because the Government failed to appeal these rulings, the finding was binding
at resentencing and precluded an actual loss enhancement to his sentence.
      As an initial matter we note that both of these objections were limited to
the calculation of actual loss suffered by the insurers, and did not involve the
actual loss to the lender, Portigon. Indeed, calculation of actual loss suffered
by Portigon was not added to the PSR until after Bazemore made these
objections. There is no indication in the record that Bazemore filed an objection


                                         7
                                 No. 15-10805

to the actual loss calculation as to Portigon during his first sentencing. During
the first sentencing, the district court did not make any factual findings
regarding the lender’s actual loss for the purposes of a loss enhancement and
therefore the law of the case doctrine did not dictate any outcome with respect
to this loss at resentencing.
      With respect to the district court’s findings on the insurers’ actual loss,
we conclude that neither the law of the case doctrine nor the mandate rule
required a finding of zero actual loss by the insurers at resentencing. The
district court, in responding to Bazemore’s objections, did not make a factual
finding on the actual loss amount. Rather, it made a factual finding regarding
how the insurer’s actual loss could be calculated. The district court’s statement
that it would “not include this figure [i.e., the gains to Bazemore from
commissions] in the [insurers’] loss amount” did not involve an expansive
finding that there was no actual loss at all. Furthermore, as the district court
recognized, sustaining Bazemore’s objection to the actual loss calculation
method “d[id] not change the loss amount” because the district court ultimately
used intended loss, not actual loss, to calculate the loss enhancement. See U.S.
Sentencing Guidelines Manual § 2B.1.1 cmt. n.3(A) (providing that the greater
of actual loss or intended loss should be used for determining the loss
enhancement). And the district court did not address Bazemore’s reasonable
foreseeability objection during the hearing, which was the only objection where
Bazemore contended that actual loss should be zero. In the cursory discussion
that Bazemore cites, the district court did not make any implicit finding
regarding the amount of actual loss for the loss enhancement, let alone a
finding of zero actual loss. Thus, the district court was not bound by the law
of the case doctrine or the mandate rule at resentencing to find zero actual loss
to the insurers for purposes of the loss enhancement. See 

                                       8
                                  No. 15-10805

(noting that “the [mandate] rule bars litigation of issues decided by the district
court” (emphasis added)); see also Wright et 

, at § 4478 (“Actual
decision of an issue is required to establish the law of the case. Law of the case
does not reach a matter that was not decided.” (footnote omitted)).
B. Bazemore’s actual loss enhancement
      The Guidelines provide for a tiered enhancement to the base offense level
depending on the dollar amount of loss relating to a defendant’s offense. U.S.
Sentencing Guidelines Manual § 2B1.1(b). The Guidelines define “loss” as “the
greater of actual loss or intended loss.” 

cmt. n.3(A). “Actual loss”
is “the reasonably foreseeable pecuniary harm that resulted from the offense.”

cmt. n.3(A)(i). To be “reasonably foreseeable,” the harm must be
“pecuniary harm that the defendant knew or, under the circumstances,
reasonably should have known, was a potential result of the offense.” 

cmt. n.3(A)(iii)–(iv). In calculating the amount of loss for purposes of
the enhancement, the district court “need only make a reasonable estimate of
the loss.” 

cmt. n.3(C). We recognize that the district court is “in a
unique position to assess the evidence and estimate the loss based upon that
evidence,” 

we give the district court’s loss determination
“wide latitude.” United States v. Taylor, 

, 564 (5th Cir. 2009) (per
curiam) (quoting United States v. Brewer, 

, 1145 (5th Cir. 1995)).
      Bazemore challenges several aspects of the district court’s actual loss
determination. We review “the district court’s method of determining the
amount of loss, as well as its interpretation of the meaning of a sentencing
guideline, de novo.” United States v. Harris, 

, 601 (5th Cir. 2016)
(quoting United States v. Nelson, 

, 520 (5th Cir. 2013)). After
determining that the method of calculation is “legally acceptable,” United
States v. Olis, 

, 545 (5th Cir. 2005), we review the embedded


                                        9
                                    No. 15-10805

factual findings—such as whether the loss amount was reasonably foreseeable
or resulted from the offense—for clear error, United States v. Brown, 

, 341 (5th Cir. 2013); United States v. Valdez, 

, 696 (5th Cir.
2013). “We will uphold a district court’s factual finding on clear error review
so long as the enhancement is plausible in light of the record as a whole.”
United States v. Caldwell, 

, 290 (5th Cir. 2006).
      Bazemore alleges the district court committed three distinct errors in
calculating his loss enhancement: (1) the district court failed to consider
Portigon’s transfer of its interest to EAA; (2) the district court failed to include
potential profits to the lender from the loans on currently active policies; and
(3) the district court erroneously found that the losses to the insurers and the
lender were reasonably foreseeable harms that resulted from Bazemore’s
fraud. In response the Government argues that any challenge to the actual
loss suffered by the lender was waived and that the challenges to the actual
loss suffered by the insurers lack merit. 6 We first address the issue of waiver
and then proceed to the merits of each of the errors alleged by Bazemore.
      1. Waiver
      As an initial matter, the Government argues that Bazemore waived any
challenge to the actual loss suffered by the lender because he challenged only
the loss calculation as to the insurers in his original appeal. On remand the
district court may not consider issues “which could have been brought in the
original appeal” but were not. United States v. Griffith, 

, 610 (5th
Cir. 2008) (emphasis omitted) (quoting 

). At the first
sentencing the district court adopted the factual findings of the PSR, which
included a finding of a $1,747,962 actual loss to Portigon for the purposes of



      6The Government also argues that any error in calculating the loss enhancement was
harmless. But because we find no such error, we do not address this argument further.

                                          10
                                  No. 15-10805

restitution. Bazemore did not appeal this specific finding of actual loss; he
appealed only the district court’s findings of actual loss relating to the insurers.
See Bazemore I, 608 F. App’x at 216–17 (vacating the district court’s order of
restitution, concluding that the district court had improperly calculated the
actual loss suffered by insurers). But we disagree that this resulted in waiver.
While Bazemore did not challenge the Portigon actual loss finding, the PSR
expressly recognized that finding was “for restitution purposes.” The “‘loss’ for
determining the offense level is different from the ‘loss’ for restitution
purposes.”   Robert W. Haines, Jr., et al., Federal Sentencing Guidelines
Handbook 1556–57 (2015–2016 ed.) [hereinafter Sentencing Handbook]; see
also United States v. Patterson, 

, 1327 (11th Cir. 2010) (“While
the Mandatory Victims Restitution Act . . . requires a judge to order restitution
to compensate the full amount of each victim’s losses . . . it does not require
restitution to match the loss figure used for sentencing. Indeed, the amounts
of loss and restitution can and do differ.” (citations omitted)). And the district
court did not make any findings during the first sentencing on actual loss
suffered by Portigon for the purposes of determining the loss enhancement.
Bazemore therefore had “no reason” to challenge the actual loss suffered by the
lender in his original appeal. 

. His failure to do so did not
result in waiver.
   2. Portigon’s transfer of interest to EAA
   Bazemore contends that the district court erred at resentencing in finding
that the lender suffered an actual loss. He notes that after his first sentencing,
EAA purchased Portigon’s interest in the policies and argues that this sale
demonstrates that Portigon was fully reimbursed for its loans to Bazemore and
therefore suffered no actual loss. He also contends that EAA did not suffer any
harm from his scheme—and therefore had no actual loss—because it was fully


                                        11
                                 No. 15-10805

informed about the fraudulent nature of the loans when it purchased the
interest in the policies from Portigon. He disagrees with the district court’s
finding that actual loss is determined at the time of the first sentencing, and
instead contends that because Portigon had recouped all of its losses by his
resentencing and EAA was not harmed, there was no actual loss to the lender.
Finally he argues that Portigon’s subsequent sale of its interest in the policies
provides the best estimate of the policies’ value and the loss enhancement
should therefore reflect the sale because that value existed at the time of the
first sentencing.
      We conclude that the district court did not err when it calculated the
actual loss caused by Bazemore’s scheme based on the time of his first
sentencing. The Guidelines provision containing the tiered enhancement for
actual loss, U.S. Sentencing Guidelines Manual § 2B1.1, “contains no general
time of measurement rule.”       Sentencing 

.         The
Guidelines do contain some specific time of measurement rules for applying
certain credits against a loss amount, see, e.g., U.S. Sentencing Guidelines
Manual § 2B1.1 cmt. n.3(E)(i)–(ii), but lack any direction as to “when [actual]
loss should be measured,” Sentencing 

. Because no
general time of measurement rule exists, the district court relied on our
decision in United States v. Goss, 

(5th Cir. 2008), to find that
actual loss should be measured at the time of the first sentencing. In Goss,
this court explained that “focusing on the [fair market value of the collateral]
at the time of the initial sentencing best comports with the [G]uidelines’ plain
language.” 

(emphasis added); see also U.S. Sentencing Guidelines
Manual § 2B1.1 cmt. n.3(E)(ii) (providing that the value of collateral is
measured “at the time of sentencing”). Goss reasoned that first sentencing
should be the time of measurement because a defendant “should be neither


                                       12
                                  No. 15-10805

penalized nor rewarded for whatever effects the time spent on appeal . . . may
have on the fair market value of the collateral.” 

.
      While the court in Goss dealt with valuation of collateral for purposes of
determining the credits against loss, the court’s concern about the potential
effects of changes in market value is just as relevant to the determination of
actual loss. If the actual loss amount resulting from Bazemore’s insurance
fraud scheme was not determined at the time of the first sentencing, Bazemore
would be (fortuitously) “rewarded” with a reduced actual loss amount for no
other reason than “the time spent on appeal.” 

the court
in Goss did not establish a “blanket rule,” 

actual loss
amount at the time of the first sentencing does not run contrary to any
provision in the Guidelines, see Sentencing 

(noting
that “the current [G]uideline contains no general time of measurement rule”).
Thus, the district court’s method of calculating actual loss based on the time of
the first sentencing is “legally acceptable.” 

. Bazemore’s
argument that the sale of the collateral to EAA provides insight into the actual
loss amount at the time of the first sentencing is therefore unavailing because
that sale occurred subsequent to the first sentencing. The district court
therefore did not err when it declined to factor Portigon’s sale of its portfolio to
EAA into its actual loss calculation and instead calculated the actual loss
amount based on the time of the first sentencing.
      3. Profits from active policy loans
      Bazemore next argues that the district court erred by failing to offset
losses to the lender with any profits that may arise from loans on policies that
are still active. The district court, in calculating actual loss on the loans held
by the lender, only included the actual loss for the nine loans related to inactive
life insurance policies.    The district court excluded from the actual loss


                                        13
                                   No. 15-10805

calculation any actual loss for the four loans on active policies, i.e., the policies
on which the lender was still paying premiums and could recover the loan
proceeds “if and when the death benefit [on the policy] is paid.” While the
district court did not explain why those active polices were excluded from the
calculation, the court did note that the Government had excluded the active
policy loans from its proposed actual loss calculation as “speculative.”
      We conclude that the district court did not err by failing to include these
potential profits in its actual loss calculation for the lender. The Guidelines
require only that the district court make a “reasonable estimate of the loss”
based on the information available to it. U.S. Sentencing Guidelines Manual
§ 2B1.1 cmt. n.3(C) (emphasis added). Determining what future gains are
likely to be earned on active policy loans, or whether any gains are likely at all,
is a necessarily speculative task. Whereas the actual loss for inactive policy
loans can be calculated with reasonable precision and certainty, gains for
active policy loans, if any, are difficult to calculate and likely to change in the
future.   The Second Circuit recently addressed how to calculate similarly
uncertain loan outcomes in United States v. Binday, 

(2d Cir.
2015).    In Binday the district court calculated actual loss from a similar
insurance scheme based on the commissions and death benefits paid under the
STOLI policies, reduced by the premiums received by insurers on policies
where “the outcome [was] known,” i.e., the actual loss figure did not include
any active policies “except to consider the commission defendants received on
those policies.” 

On appeal, the Second Circuit held that this method
of calculating actual loss was reasonable because the “‘actual’ gain or loss”
could be calculated only for “[t]he policies for which the results [we]re already
known,” not for active policies where gains or losses were a mere “possibility.”

                                         14
                                 No. 15-10805

      Similarly, the only actual gains or losses from Bazemore’s scheme which
are known with reasonable certainty are the losses related to the inactive
policy loans. While Bazemore contends that the lender could potentially profit
from the active policy loans because the death benefits could be greater than
the premiums paid, Bazemore has provided little support for this contention
beyond his own speculation. And, as the Government correctly notes, the
evidence in the record speaks only to potential future recoupment of the loan
proceeds plus interest and fees, not of potential profit. For example the FBI
report summarizing an interview with Portigon representatives explains that
“Portigon has a secured interest in the policies and could recover the loan
proceeds if and when the death benefit is paid.” Another report states that
Portigon representatives acknowledge that Portigon has a security interest in
active policies involved in the schemes but cautions “it is uncertain whether
making the premium payments through maturity will be beneficial because it
is uncertain whether the insurance company will honor these policies, due to
[Bazemore’s] fraudulent actions.”      Thus, the potential profits urged by
Bazemore on the active policy loans appear to be, at best, only a “possibility.”

. We conclude that the district court’s decision to base
its actual loss calculation solely on the inactive policy loans was a “reasonable
estimate of the loss” based on the information available to the court. U.S.
Sentencing Guidelines Manual § 2B1.1 cmt. n.3(C).
      4. Reasonable foreseeability of the actual losses
      Finally, Bazemore contends that the district court erred in finding that
actual losses to the insurers and the lender were a reasonably foreseeable
result of Bazemore’s scheme.     The Guidelines define “actual loss” as “the
reasonably foreseeable pecuniary harm that resulted from the offense.” 

cmt. n.3(A)(i). This definition “incorporates [a] causation standard


                                       15
                                       No. 15-10805

that, at a minimum, requires factual causation . . . and provides a rule for legal
causation.”     

(alteration in original) (quoting U.S.
Sentencing Guidelines Manual Supp. to App. C Amendment 617 (Nov. 1,
2001)).     Accordingly, a defendant’s sentence should be based only on
“losses . . . caused directly by the offense conduct.” 

“District courts
must take a ‘realistic, economic approach to determine what losses the
defendant truly caused or intended to cause.’” 

States v. W.
Coast Aluminum Heat Treating Co., 

, 991 (9th Cir. 2001)).
       Bazemore contends that any losses to the insurers were not caused by
his fraudulent scheme but rather resulted from the insurers’ independent
decisions to refund the premiums. Noting that only some of the insurers chose
to refund premiums and that the lender chose to continue paying premiums on
some policies, Bazemore asserts that any actual losses on policies were the
result of the insurers’ independent, erroneous determinations that the policies
were not profitable—namely, “miscalculations regarding the applicants’ likely
life-span.” In addition he argues that he did not cause any loss to EAA because
the causal chain was “broken” by EAA’s “fully informed decision . . . to buy the
policies” from Portigon. Therefore he concludes that he did not cause any
actual loss to either the insurers or the lender. 7
       Reviewing the record as a whole, we conclude that the district court did
not err in finding that losses from the insurance policies were a reasonably
foreseeable       harm       that      resulted      from       Bazemore’s        fraudulent


       7  The Government contends that this argument is foreclosed by the law of the case
doctrine, which required the district court to apply the actual loss formula described in
Bazemore I for losses on rescinded policies. See 608 F. App’x at 217 (“The actual loss on a
rescinded STOLI policy is the commission the insurer paid to Bazemore less any premium
payments it retained.”). But this argument is inapposite because here Bazemore is not
challenging the method of calculating actual loss to the insurers; rather, he is challenging the
district court’s factual finding that he caused the loss to the insurers.


                                              16
                                    No. 15-10805

misrepresentations. The record demonstrates that insurers would not have
issued the insurance policies but for Bazemore’s misrepresentations of the
applicant’s net worth and intention to transfer the policy to an investor. See
Bazemore I, 608 F. App’x at 209 (“The insurers that issued policies to
Bazemore’s applicants would, without exception, deny life insurance policies to
applicants that intended from the outset to transfer the policy to a third
party.”); Sentencing 

(“[A] loss that ‘resulted from’ an
offense is one that would not have occurred but for the occurrence of the
offense.”). And Bazemore knew that, as the agent responsible for the sale, he
would receive a commission on each issued policy roughly equivalent to the
amount of the first year’s premium.            

Bazemore made
fraudulent misrepresentations to the insurers regarding the net worth and
intentions of the applicant, he “knew or, under the circumstances, reasonably
should have known” that a potential result was monetary harm to the
insurers—namely, the commissions they paid to Bazemore.                    See U.S.
Sentencing      Guidelines    § 2B1.1   cmt.    n.3(A)(iv)   (defining   “reasonably
foreseeable pecuniary harm”).        Furthermore the record indicates that life
insurance policies for high net-worth individuals are often used as investment
vehicles or financial instruments, thus it was foreseeable that the lender could
transfer its interest in these policies to another party. Because the policies
were the sole purpose behind (and the collateral supporting) the loans by the
lender, Bazemore knew or reasonably should have known that losses to the
lender would result from his fraudulent misrepresentations on the policy
applications.         See    U.S.   Sentencing       Guidelines   Manual     § 2B1.1
cmt. n.3(A)(i)(iv).
      We reject Bazemore’s argument to the contrary. In particular, he fails
to provide any meaningful support in the record for his argument that the


                                         17
                                  No. 15-10805

divergent outcomes of some of the policies and loans show that any loss
resulted from the victims’ own “miscalculations regarding the applicants’ likely
life-span.” Standing alone, Bazemore’s argument is insufficient to conclude
that the district court did not make a plausible factual finding of causation
based on the record as a whole. See 

. Thus, the
district court did not commit clear error in finding that the actual losses to the
insurers and the lender were “reasonably foreseeable pecuniary harm that
resulted from [Bazemore’s] offense.”        U.S. Sentencing Guidelines Manual
§ 2B1.1 cmt. n.3(A)(i).
            III.   VIOLATION OF THE PROFFER AGREEMENT
      The Guidelines generally prohibit information provided by the defendant
“pursuant to” a cooperation agreement from being used to determine the
applicable sentencing range. U.S. Sentencing Guidelines Manual § 1B1.8(a).
Bazemore shared the details of his insurance fraud scheme with an FBI agent
while he was covered by an agreement to proffer evidence in an unrelated
securities fraud case. Bazemore contends that these statements to the FBI
agent were covered by his proffer agreement and the district court, via
imposition of the loss enhancement, improperly used information gleaned from
these statements to determine his Guidelines sentencing range. We review de
novo whether the Government has breached a proffer agreement. See United
States v. Gonzalez, 

, 886 (5th Cir. 2002).
      Here we do not reach the merits of Bazemore’s argument because we
conclude that he waived it by failing to raise it in his original appeal. During
Bazemore’s first sentencing (like his resentencing), the district court used
evidence that the Government obtained following Bazemore’s statements to
the FBI in order to calculate the appropriate loss enhancement to his sentence.
In response, Bazemore objected that his proffer agreement precluded the


                                       18
                                       No. 15-10805

Government from using this information, but the district court expressly
overruled that objection. Bazemore failed to challenge that ruling when he
appealed his first sentence. His failure to do so bars him from making this
argument now.
       Bazemore attempts to circumvent waiver by arguing that his original
appeal involved an objection to the intended loss amount rather than the actual
loss amount and that “[t]he mandate rule does not require parties to raise all
issues that are similar or analogous to claims that might be made on remand.”
This distinction is unpersuasive. As clearly stated in the Guidelines, the loss
enhancement is based on the “greater of actual loss or intended loss.” U.S.
Sentencing Guidelines Manual § 2B1.1 cmt. n.3(A).                      And as Bazemore
articulated in his objections during his first sentencing, “any enhancements
regarding loss or committing insurance fraud . . .[were] protected under . . . the
proffer agreement and cannot be used to increase Bazemore’s [G]uideline
range.” (Emphasis added). The alleged violation of the proffer agreement
affected the district court’s first loss enhancement calculation (using intended
loss) as much as it affected the court’s loss enhancement recalculation (using
actual loss) on resentencing. Therefore Bazemore had every reason to raise
this alleged error when appealing his first sentence. Bazemore has waived this
proffer agreement argument by failing to raise it in his original appeal. 8 See

.
                            IV.    APPRENDI CHALLENGE
       It is well established that any fact—other than a prior conviction—that
increases a defendant’s maximum sentence “must be submitted to a jury, and


       8 While Bazemore alleges that a similar issue underlies the district court’s application
of a role enhancement to his sentence, Bazemore also failed to raise any objection to that
enhancement during the original appeal and thus it is also waived. Bazemore I, 608 F. App’x
at 213.

                                             19
                                    No. 15-10805

proved beyond a reasonable doubt.” Apprendi v. New Jersey, 

, 490
(2000). Bazemore argues that the district court violated this principle by
making a factual finding—actual loss amount—that increased his maximum
sentence. Although recognizing that this court has previously rejected this
argument in United States v. Tuma, 

, 693 (5th Cir. 2013),
Bazemore suggests that the Supreme Court’s recent decision in Hurst v.
Florida, 

(2016), which was decided after his original appeal,
calls Tuma into question. We review alleged violations of Apprendi de novo.
See 

.
      To the extent that Bazemore is arguing that Hurst provides an
intervening change in law that excuses his failure to raise this issue on his
original appeal, see 

, this argument is unavailing
because Hurst’s holding is inapplicable.           Hurst held unconstitutional a
statutory scheme in which “the maximum sentence a capital felon [could]
receive on the basis of the [jury] conviction alone [was] life imprisonment” and
the felon could receive a death sentence only if the court made additional
findings at a subsequent sentencing 

. Hurst
therefore applies only to statutory schemes in which judge-made findings
increase the maximum sentence that a defendant can receive. 

In
contrast, here the district court’s actual loss findings did not increase the
maximum punishment that Bazemore could receive; it sentenced him to 188
months’ imprisonment, less than the 20 year statutory maximum for a mail
fraud conviction. 9 See 18 U.S.C. § 1341; see also 

, 693
(upholding that district court statutory maximum sentence).              This “broad
sentencing discretion, informed by judicial factfinding, does not violate the



      9  The court’s actual finding did alter Bazemore’s advisory Guidelines sentencing
range, but it did not change the maximum punishment allowable by law.

                                          20
                                    No. 15-10805

Sixth Amendment.” 

; see also United States v. Mares,

, 519 (5th Cir. 2005) (“The sentencing judge is entitled to find by
a preponderance of the evidence all the facts relevant to the determination of
a Guideline sentencing range and all facts relevant to the determination of a
non-Guidelines sentence.”).
      Because Hurst does not provide a new basis for challenging his sentence,
Bazemore waived this challenge to his sentence by failing to raise it in his
original appeal. This Apprendi argument was relevant to Bazemore’s prior
appeal; at his first sentencing, the district court used its own findings of fact to
determine the loss enhancement to Bazemore’s sentence. It makes no material
difference that the finding of fact at the first sentencing was on the intended
loss amount whereas the resentencing finding was on the actual loss amount.
See U.S. Sentencing Guidelines Manual § 2B1.1 cmt. n.3(A) (stating that the
loss enhancement is based on the “greater of actual loss or intended loss”).
Thus, whether judicial factfinding of a loss enhancement violated Apprendi
“could have been brought in the original appeal,” 

(emphasis omitted) (quoting 

), and is therefore waived in
this subsequent appeal, 

                               V.     CONCLUSION
      For the foregoing reasons we AFFIRM Bazemore’s sentence.




                                        21