TSP Talk was never intended to be a "watchdog" organization, but that does not mean we can't analyze this situation.
The press release indicated they were concerned about 3000 folks who are rebalancing daily. Given 20 trading days a month, a two (or four roundtrip) IFT limit essentially represents about 90% or 80% reduction, of these "problem children" ability to IFT. A 60% reduction (4 trades into stocks and unlimted trades to the G or 8 trades total limit) seems like a less dramatic solution.
The minutes clearly show that the board recognizes a need for the system to be mega crash proof (in the event of a catastrophic event - i.e. something happens in the world, like a September 11th, and everyone runs for the safety of the G). In order to achieve that capability, the daily IFTs are irrelevant. So the issue is not storage capacity or hardware.
The minutes indicate that the problem seems to stem from the FV process when it fails to accurately predict (using the FV) the buying price (the purchase from Barclay's is executed the next morning). By the admission of the minutes, this seems to occurr when volatility is at it's steepest - which is also what causes unusally large number of exchanges. So when the FV process is most likely to be inaccurate is also when the highest number of IFTs is likely to occurr - therefore, limiting the IFT's may have little to no impact on this problem. A uesful argument - which requries quite a bit of statistics to show - and if the board has done the math then good for them - but if they haven't done the math, then shame on them and they need to evaluate the effect their solution would really have before they implement this unproven policy.).
Another solution, they may have not considered, would be for TSP to postpone the pricing of funds until AFTER the I-fund shares were purchased from Barclay's the following morning - which means that TSP would not post the numbers until somewhere around 1100 EST. The upside of this though, may be that they could extend the deadline. Let's say they pushed it out to 1500EST. Would you rather have things work they way they are now or do it this way? IMHO - this is a much better solution - becasue we are all still going to be able to guess the closing prices just as we do now - except that we might be a couple pennies off, because of the onsies/twosies of large movements within the funds themselves influencing the price. Plus, we would not have to make decisions about our IFT until after noon - does it really matter to the timer, if they don't get the actual numbers until the next morning? That's rhetorical - the answer is No.
I would appreciate y'all's thoughts.
