The $36,500 Round-Trip — Checks 1031 & 1038

🛑 CORRECTED 2026-06-29 (supersedes the "THREE round-trips" reading below). Landon's review of the Bates-stamped disclosure checks establishes only TWO round-trips, not three: - 2017 — $36,500 round-trip, JHF check in JWI's hand dated 11/27/2017 (Bates JWI000467–JWI000468). (= the "ck#1017" template.) - 2018 — $36,500 round-trip, JHF check in JWI's hand dated 12/31/2018, cleared 1/4/2019 (Bates JWI000450–JWI000451). (= "ck#1031".) - 2019, 2021, 2022 — NO rent paid at all (no check in or out). The earlier "FY2019 ck#1038" round-trip is withdrawn — the Bates disclosures show no 2019 check either direction; the apparent third instance was the 12/31/2018 check that cleared on 1/4/2019, counted twice in the spreadsheet. - 2016 — a $34,500 "rent" check dated 11/28/2016 (Bates JWI001401) was produced as proof but never deposited (operating account closed Aug 2016) — a phantom payment, not a round-trip. - 2023 — rent paid ~4 months late (April 2024), only after Bill's death prompted questions. The referral (2026.06.13 Imhof Criminal Allegations v2) now reflects this two-round-trip + phantom + non-payment record. The signature analysis (genuine 2017 "Lewis C. Frenzl TTEE" vs. questioned 2018) is unchanged.

⚠ Disambiguation: bank check #1031 (this $36,500 paper check) is NOT the IRS § 1031 like-kind exchange (the 2023 Condon/Harman land deal). Unrelated; the number is a coincidence. Keep them in separate lanes.

📊 Update from the account ledger (2026.06.15 JHF Trust Account Ledger — Chronological Backbone): there are THREE round-trips, not two — the maneuver ran in FY2017 (ck#1017), FY2018 (ck#1031), and FY2019 (ck#1038), then it stopped. Three consecutive years strengthens the § 18-4-401(4)(b) course-of-conduct argument. ✅ FY2017 check pulled (2026.06.16): it is CHECK #1017 — $36,500.00, written 11-27-2017, payable to Lewis Frenzl, cleared 11/28/2017, matched same-day by a $36,500.00 DDA Regular Deposit (slip "John Frenzl Trust," acct 5024132). Statement 11/13/17–12/08/17: balance $25,817.87 → +$36,500 → −$36,500 → $25,817.87. Net zero. Confirmed. Source: SGR/2024.11.01 FARMERS STATE BANK/2017 STMTS 5024132/2017.11.xx 5024132.pdf p.2. 🔑 Why ck#1017 matters to the forgery question: it is the genuine in-series template. Same payee, same $36,500, same maneuver, same account — but signed "Lewis _C._ Frenzl TTEE" (middle initial and trustee designation, both clearly present — confirmed on the 600-DPI crop). One year later ck#1031 drops both the middle "C." and the "TTEE," and gains a "Justin Imhof" notation. The signature degraded in lockstep with Justin's appearance on the instrument. ck#1017 is therefore the single best known-genuine exemplar to put beside the questioned ck#1031 — they are the same recurring check one year apart.

Checks 1031 and 1038 are two of three separate instances of the same maneuver, a year apart — not a single check mis-numbered. The prior ICAv2 draft "corrected" 1031 into 1038 as if 1031 were an error. That was wrong. Both checks exist, both are $36,500.00, both are payable to Lewis Frenzl, and each cleared the same day as a matching $36,500 deposit back into the Trust.

Side-by-side

The pattern: at each year-end (FY2018 and FY2019) a $36,500 check left the Trust to Lewis Frenzl and an equal $36,500 came right back into the Trust the same day. The money never actually went anywhere — it was a paper round-trip. This is the "laundering back into the LCF account" maneuver, run two years running.

O-1 reframed (per Landon): 1031 and 1038 are separate issues. 1031 (Dec 2018) is the forged-signature instance; 1038 (Dec 2019) is the same-pattern laundering instance the following year. Do not discredit 1031.

Forgery-review links — Check 1031 (for Landon's manual confirmation)

I cropped the scan so you don't have to dig for it. Open these directly:

• Full check 1031 + body + signature: C:\Users\lando\CHECK_1031_2018.12.31_36500_FORGERY-REVIEW.png
• Signature close-up: C:\Users\lando\CHECK_1031_SIGNATURE_closeup.png
• Matching $36,500 deposit slip (1/4/2019): C:\Users\lando\CHECK_1031_matching_36500_DEPOSIT_2019.01.04.png
• Full statement page (3 checks + deposit): C:\Users\lando\check1031_2018-12_p2_full.png
• Original source PDF: CO/BarFZ/SGR/2024.11.01 FARMERS STATE BANK/2018 STMTS 5024132/2018.12.xx 5024132.pdf (p. 2)

Genuine in-series template — Check 1017 (FY2017):

• Full ck#1017 (genuine exemplar): C:\Users\lando\CHECK_1017_2017.11.27_36500_GENUINE-EXEMPLAR.png (also _FINAL/EXHIBITS/A_36500_Checks/)
• Signature close-up (shows "Lewis _C._ Frenzl TTEE"): C:\Users\lando\CHECK_1017_SIGNATURE_closeup.png
• Original source PDF: CO/BarFZ/SGR/2024.11.01 FARMERS STATE BANK/2017 STMTS 5024132/2017.11.xx 5024132.pdf (p. 2)
• Put 1017 beside 1031: it is the same check, one year earlier, signed correctly. The strongest single comparison we have.

What I see on the signature close-up (confirm manually): the main signature line carries a cursive "Lewis Frenzl"; below the line and to the left, near the "1031" check number, there is a second, smaller cursive notation that reads "Justin Imhof." Two independent questions follow:

1. Is the "Lewis Frenzl" signature genuine? ← this is the forgery question. (O-4: own forgery-probability analysis.)
2. Why is "Justin Imhof" written on the signature block of a 2018 trust check when Justin held no fiduciary title until 2/15/2023? Even if Lewis's signature is genuine, Justin's presence on the instrument is independently significant.

Preliminary in-house signature comparison (ck#1031 vs. known LCF) — NOT expert opinion

Exhibit built: C:\Users\lando\SIGNATURE_COMPARISON_ck1031_vs_LCF.png (also staged in _FINAL/EXHIBITS/A_36500_Checks/). It stacks the questioned ck#1031 signature block over four known-genuine Lewis C. Frenzl signatures from recorded instruments (WaCo REC#854431 [2012], #857958 [2013], #857959 [2013], #861774 [2015]). The FSB signature card (LCF TJI JWI Signature Panel.png) supplies a fifth genuine Lewis + genuine Justin exemplar.

Observations (mine — to be confirmed/refuted by a qualified document examiner):

1. Middle initial. All five genuine exemplars sign "Lewis _C._ Frenzl" with a clear middle "C." The questioned ck#1031 reads "Lewis Frenzl" — the middle initial appears absent.
2. "TTEE" designation. Lewis's genuine trustee signings (854431, 857958, the FSB card) append "TTEE." ck#1031 — itself a Trust check — lacks it. (Caveat: two genuine exemplars, 857959 and 861774, also omit TTEE, so this is supporting, not conclusive.)
3. "Justin Imhof" on the block. The strongest single feature: Justin's name is written on ck#1031's signature line — and the FSB card confirms Justin as an authorized signer, so his hand on the check is consistent with him supplying the second lock-out signature.

Preliminary read: the questioned signature shows class-characteristic differences from the genuine set consistent with the forgery/JWI-execution hypothesis. This is not a forensic conclusion — a single scanned image at statement resolution cannot establish forgery; that requires a qualified examiner working from the original instrument. Treat as "flagged, supported, pending expert."

Tools for the signature work (recommended)

• Already installed & free: OpenCV + scikit-image — align two signatures (ORB/SIFT feature match + homography), then overlay and compute SSIM/difference maps to quantify stroke divergence; Pillow for cropping/compositing (used to build the exhibit above).
• Manual overlay: GIMP (free) or Photoshop — transparency-overlay the questioned sig over a genuine one to compare proportion, slant, pen-lifts. (We already have .psd exemplars in LCF Signatures/.)
• **For a similarity score: a Siamese signature-verification model (SigNet) pretrained on the CEDAR/GPDS signature datasets — feed genuine exemplars + the questioned sig, get a genuine/forged similarity probability. I can stand this up locally (PyTorch) if you want a numeric in-house estimate; install ask:** pip install torch torchvision (CPU build is fine).
• Highest-resolution input: re-render ck#1031 from the source PDF at 600 DPI and crop just the signature for any of the above (the current crop is 400 DPI).
• The human expert (forensic document examiner / QDE) remains the authority for any released opinion; the above is only for our own pre-release confidence.

Handwriting attributions on the surrounding checks (confirmed by Landon)

From the December 2019 statement page (2020_…pdf p.2):

The contrast is the point. The legitimate scholarship checks (1033/1035/1037) are in Krening's hand and signed by Lewis. The $36,500 round-trip check (1038) body is in JWI's hand. The two deposit slips are in two different hands — consistent with two different tellers at Farmers State Bank of Akron taking the deposits in person. A handwriting examiner can use Krening's known scholarship-check exemplars and JWI's known exemplars as the comparison baselines.

Why this matters to the charges

• Forgery (§ 18-5-102) — if the "Lewis Frenzl" signature on 1031 is not genuine, this is a stand-alone forgery count, separate from 1038.
• Unauthorized control / theft (§ 18-4-401) — the $36,500 body in JWI's hand on 1038 (and "Justin Imhof" on 1031's signature block), while JWI held no title, is an unauthorized exercise of control over Trust funds.
• Course of conduct (§ 18-4-401(4)(b)) — the same $36,500 round-trip, two years running is itself evidence of "one scheme or course of conduct" against the same person/entity, supporting aggregation (see 2026.06.15 Section 17 — How 560ac Reached Justin and the ICAv2 Count 2 rework).

Open / to-do

• ☐ Landon: manually confirm the "Lewis Frenzl" signature and the "Justin Imhof" notation on 1031 (links above).
• ☐ O-4: commission our own forgery-probability analysis of 1031 (and 1038) — lower felony priority, kept on the list.
• ☑ FY2017 pulled (2026.06.16): ck#1017, genuine "Lewis C. Frenzl TTEE" signature — the in-series template. Crops staged.
• ☐ Pull the FY2020 year-end from the statements to confirm the round-trip stopped after 1038 (the ledger shows it did; confirm visually on the check page).
• ☐ Re-render ck#1017 + ck#1031 signatures at matched 600 DPI and overlay (OpenCV/SSIM) — same-check, one-year-apart comparison is the cleanest signal we have.