Jenny Johnson's $150 legacy cost per fund transaction is 100 times the figure she gave in January

Johnson's two accounts of the same comparison are exactly 100 times apart on the legacy side ^[15]. The TOKEN2049 version, $1.13 against roughly $150 a transaction ^[3], works out to about 133 to 1 ^[16]. The January version is odder, and better for the chain. On those figures the legacy system would have spent about $75,000 on the 50,000-transaction batch ^[17], while the Stellar run cost roughly 0.002 cents a transaction ^[18]. That is a gap of around 66,000 to 1 ^[19].

A figure exactly 100 times the January one is most simply explained by a decimal point lost somewhere between the stage at the Oct. 8 conference ^[2] and the report. The other explanations are a cost basis Franklin has not published, or a January study the firm no longer stands behind. crypto.news, which set the two statements side by side, said the $150 should be treated as unverified unless Franklin supplies a separate calculation or methodology ^[4].

Whichever it is, the $1.13 is what it cost to run the batch on Stellar ^[12], and the chain holds only part of BENJI's record. The SEC staff letter dated Aug. 12 says Franklin Templeton Investor Services keeps an internal system of private shareholder information next to the anonymous on-chain transaction data ^[6]. The two are linked in real time to form the official shareholder file ^[6]. Franklin has not published an all-in cost per transaction that includes that internal system.

The structural argument holds up better than the cost one. Franklin's research defines a digital twin as a token linked to an ownership record kept offchain ^[13]. Its own model puts the official transaction and ownership data inside a blockchain-integrated system ^[14]. That gives an allocator two questions for any tokenized fund: does the official shareholder file depend on the chain, and who can change an entry? BENJI answers yes to the first and its transfer agent to the second ^[7]. A twin, on Franklin's definition, answers no to the first ^[13].

Franklin runs BENJI on nine public networks ^[10]. It has also used the tokens as deal currency (the April 250 Digital agreement included BENJI tokens as payment consideration before it completed in June) ^[11].

In my view the record distinction survives scrutiny, and the cost claim in its $150 form does not yet. The counter-case is that the transfer agent's override powers keep BENJI closer to the twins than Johnson allows ^[7]. An all-in per-transaction cost from Franklin, internal system included, that still came in well below the $1.50 legacy figure would settle the cost half in her favor ^[12].