Quote from: BitBits on October 12, 2014, 05:57:58 PM

Quote from: QwertyCore on October 12, 2014, 03:06:10 PM

Quote from: BitBits on October 12, 2014, 01:30:09 AM

Hello Phil,

I've personally seen good use of FRR feature in the past, but these days I would agree with noggin-scratcher regarding the issue of "an enormous wall of offers all placed at the single FRR" and the resulting huge spikes (like today, to 0.7%!!!) when this wall is gone. I do not think however that FRR feature needs to be eliminated entirely.

To your list of simpler solution suggestions, I would propose to limit the total volume of FRR offers to a certain percentage of total offers on the order book. I don't know what percentage allocated to FRR will be optimal, but I guess you could start from some high value and go down over time, to ensure smooth transition.

It basically comes to the (optional) requirement for users to specify the rate when they are placing FRR offer. We already have a box for this entry, but currently it gets ignored by the system, if user clicks on FRR option, so perhaps we should try to enable that.
Under simplest implementation of this feature, user's offer will be placed at FRR, or at specified by him/her rate if FRR quote is already taken.

Slight modification of the above will be the availability of additional options, when placing the swap offer, such as the following:
1. Place offer at FRR, or if not available - at user specified rate (this is the same as described above)
2. Place offer at FRR, or if not available - at equal to then current FRR rate (this can be useful, for those using auto-renewal option)
3. Place offer at FRR, or if not available - do NOT place any fixed rate offer, wait until FRR offer CAN be placed.

Note that the option 3 above is very similar to what we have right now (meaning that your offer will "stand in line" for placement at FRR). For this reason, option 3 can be made a default setting. Options 1 and 2 will create much thicker order book and prevent big spikes.

Other than the above, availability of some algos would be ideal and by the way, if you do not want to make the whole platform "overly complicated" (this is of course understandable), perhaps you can make algos available only to those who wants to use them, leaving simpler solution for the rest. I guess BFX will run into problem when people would just "enable" the algos and then annoy the hell out of support team asking basic question instead of reading some sort of a "how to" . To avoid this, a comprehensive "How to" section can be created and users wishing to use the "algo system" would have to pass a simple quiz, before it is enabled on their account for them.

What do you (and all) think?

None of these ideas will solve the problem, all they will do is hide it. ...

This was just one idea and you are pointing out at only one minor sub-point of everything, going on and on about it and saying that the whole idea cannot work. You simply missed it.

Forget FRR queuing, you are right on that, cancel that option 3, make it as following:
...
3. Place offer at FRR, or if not available - do NOT place any fixed rate offer. (send email notification).

Look, everybody agrees that huge wall at FRR is a bad thing these days, right? It was then suggested to get rid of FRR feature completely. What I suggested is to make a step "in between" and that is, to limit the maximum volume for FRR offers on order book.

I didn't mean to come across as nit-picking, I was just trying to show that all the scenarios degrade to the same state.

My main point is that I think the whole concept of the FRR is fatally flawed. It is a coalescent point for laissez faire lenders. If you allow spreads to the FRR such as FRR±x% or FRR±x you might not have a wall, but you are going to have a very steep hill around the FRR that will function the same way.

I thought about applying the same tools that the trader have (stop loss, training, etc) to swaps but swaps and trading are not the same animal. With trading, when you buy, the deal is done, you own what you traded for. The is not the same with swaps and I think why that is why the FRR exists at all.

What is causing the FRR to exist/be debated is the fact that the trader is not committed holding the swap for the entire term (and I'm not suggesting that they should). If a lender could place a swap for 5 days and know that it will be held the full 5 days, I think the FRR would be less widely used and the wall would fall. A lender would be more likely to hold out for a higher rate if that would be locked in for 30 days. As it is, a lender can put an FRR offer in and know that if the swap is closed early he still has and offer on the books at a competitive rate. It might not be a good rate, but getting some return is better than nothing.

I do not have a solution to fixing the FRR wall, but I know the wall is bad for active lenders. In my opinion the FRR should never have existed and should be removed until a better way is found. Personally, I don't think a better way exists that not favour either the active or passive lender over one another or favour the lenders or traders over each other.