... and as my understanding will solidify I would like draft summaries and recommendations in the IT lingo for development
... I can commit myself to closely watch glbse related threads ...
thus said ^^
I do have some CS background, so if you would like to translate the "shoulds" into dev-lingo, I'd be thrilled to read them and make comments.
that's my plan. to check with you if the change requests (feature requests) make sense and add to the usability of glbse for ETF
there are also some practical questions, such as:
question is: should the ETF pricing be derived from the aggregated underlying bid/ask?
should the pricing be done by the issuer or from GLBSE ? etc etc
as long ETF is a standard asset with generic buy/sell orders (and nothing else), GLBSE has no tools or rules how it could influence the pricing (it is the responsibility of the asset manager and his ability to react accordingly on market moves).
I have only a basic understanding (imagination) how pricing derived from bid/ask could be dome. such option would simplify the "minute to minute" operations right there at the server side without the need to develop a bot. it makes sense as well to leave this repetitive and error prone task to a computer. my first thought was a but ... the price would be then 'free floating' subject to all (also to sudden) swings in supply/demand and price changes. pricing by glbse would require new types of asset where the issuer would agree to that and the rules would be well known and properly supported by the software.
is not glbse too small for that yet anyway? moving prices with control of let's say 20-40% shares in 2/5 of constituents should be easy, especially if bot would react in real time and could be trapped in a position that triggers asset sell off (note to myself, thing of built in limits, conditions to check for, border conditions and user/issuer preferences p.ex.)
if nothing would change, ETF manager would be as any other asset manager, manually creating orders, by your previous note closer to managed fund than ETF and deviate from constant ratio of all constituents at all times (p.ex. great differences in price per stock, up to 15x cheaper/more expensive shares of mining companies exist. sales of the expensive stuff would occur only 1 in 15 times where the cheaper asset should be sold.)
assuming trading volume will not burst and spreads would be reasonable, how the aggregate pricing formula would look like in real world exchange, please?