I'm trying to value a portfolio of CDX's, most of which are off-the-run (currently the on the runs are S13's). For many off-the-run CDX's Markit do not provide a quote. I have also tried CMA but I'm not sure if their quotes are reliable as they can be based on historical/stale data (they show no quotes/liquidity on those CDX's). In such situation what could you do? Where could you go for a spread/price? Or just stick with CMA's and hope for the best?

QuoteOriginally posted by: d32I'm trying to value a portfolio of CDX's, most of which are off-the-run (currently the on the runs are S13's). For many off-the-run CDX's Markit do not provide a quote. I have also tried CMA but I'm not sure if their quotes are reliable as they can be based on historical/stale data (they show no quotes/liquidity on those CDX's). In such situation what could you do? Where could you go for a spread/price? Or just stick with CMA's and hope for the best?can you derive spread for not on run CDX usinng on the run CDX? Do not trust CMA data. Regulators should stress CME.

Some of the earlier off the run CDX may have different (or less) names making up the index than the newer ones, via credit event etc. Can you still imply the spread in this case?

Different series have different constituents, so you can't interpolate across them. CDX off-the-runs are quoted fairly well, several dealers quote both rolls and curves on old series. Not sure if any data service aggregates that data though.

Given there is no/little trading activity in OTRs you could mark them where you want i.e. treat them as a bespoke. At month end there is a consensus run by Markit, which generally matches any independent broker marks. So you have market prices periodically. But who knows how these month end prices are calculated?Furthermore given the illiquidity if a big dealer does not like the market price of an OTR they have on there books, they can move the market price to the prefered price.

MaxCohen - CDX IG OTRs aren't that illiquid. Really old series rarely trade, but S6 and later are fairly transparent. You can get a 2 way market in these easily. Bid-offer won't be same as the on the run of course.

d32, perhaps the "on-the-run" tranche should be considered as the most liquid series not the most recent (is this CDX9 at the moment?). Then if treating OTRs as bespokes (those OTRs for which there is no market price) you can imply the price by using the most liquid or on-the-run series and some bespoke mapping methodology.StructCred, what does d32 do when they need to value a book consisting of CDX3-13 and ITRX2-12 tranches. You need some consistent methodology to value the portfolio daily. Days when there are no market prices for a given OTR series mark-to-model, then when you observe a price (which you think is valid/transparent) mark-to-market.

QuoteOriginally posted by: d32I'm trying to value a portfolio of CDX's, most of which are off-the-run (currently the on the runs are S13's). For many off-the-run CDX's Markit do not provide a quote. I have also tried CMA but I'm not sure if their quotes are reliable as they can be based on historical/stale data (they show no quotes/liquidity on those CDX's). In such situation what could you do? Where could you go for a spread/price? Or just stick with CMA's and hope for the best?MarkIt have to make clearing settlement spreads for CDX publicly available. Check out www.markit.com/cds -- there is a box called MarkIt CDS Clearing Prices.http://www.markit.com/cds/prices/markit_ice.shtmlMark

sorry i thought d32 was talking about tranches not index.

Thank you all. mromwell, publicly Markit only provide previous day but no historical? I figured out a way to get historical CDX spread/price in their pay portion of the site. Now here is another problem.For certain off-the-run CDX they may not have spread/prices on the day I want but they might have days around it (say no spreads for 16th Oct, but have for 15th and 19th). If on both the 15th and 19th they have the same spread and not much vol. in the previous days then it's reasonable to just use that spread. But if they are different or worse have large change in spread day to day then anything goes, which is a problem. I have another question which is somewhat related: how is markit deriving their model price/spread on those OTR CDX? Especially if there are no market quotes for those days (or for many days).