Regarding <transfer> name$.utterance shadow variable value for DTMF input scenario

Hi,
 
I have a query regarding the value that need to be set for name$.utterance
shadow variable for <transfer> in scenario where transfer is terminated by
DTMF input.
 
According to VXML 2.0 specification, name$.utterance  will be set using "the
utterance text used if transfer was terminated by speech recognition input
or the DTMF result if the transfer was terminated by DTMF input; otherwise
it is undefined. ". 
 
My query is regarding the term "DTMF result". Consider the following VXML
usage:
 
    <transfer name="mycall" dest="tel:+1-555-123-4567"
               transferaudio="music.wav" connecttimeout="60s" 
               type="bridge">
       Please wait while I transfer you to the operator. Otherwise, Dial
your oustanding balance amount to continue.
       <grammar mode="dtmf" src="builtin:dtmf/number" /> 
     </transfer>

Suppose the user dials the key sequence "123*45", which results in  a
grammar match (for builltin grammar "number"), and produces an
interpretation result 123.45 (* is used as decimal point).
 
In this scenario, what should be the value for mycall$.utterance shadow
variable? Should it be 123*45 or 123.45 ?
 
Thanks
Rajesh