sorry guys just one more question relating to prepayments. would you expect housing turnover to have a higher effect on prepayments than refinancing? i always thought housing turnover is the most significant driver of prepayments, but i just recently saw some analysis that showed an interest rate differential had a higher effect (but this analysis was on the breakage cost penalty the bank receives for early prepayments, isntead of on the actual amount prepaid). just wondering what you guys would expect to see. my analysis would be relating to canadian mortgages. thanks in advance!
>>would you expect housing turnover to have a higher effect on prepayments than refinancing? I don’t know anything about Canada but you should be able to pull data on count of refi’s and the count of home sales and see how it varied over time. Pull prepayment rates in as well and compare. Interest rate is far and away the main driver of prepayment rate, but of course turnover influences it too.
thanks juventurd! ill definitely do that once i get the data, just wanted to get an idea of what was expected.
There’s a certain level of housing turnover that is just “normal” (might be higher these days because of foreclosures) as people grow families or relocate for work or whatnot. Demographics and employment affect this level. The issue with refinancing is that it is highly correlated (inversely) to interest rate changes. As mortgage interest rates drop, more people refinance. Housing turnover may change as well, since lower rates in normal times increases housing affordability, but my guess is that the effect is not as strong.
A good chunk of prepayments are due to housing turnover, but as bchad said this is more or less stable over time. When you look at an MBS you’re concerned about the movement of the prepayment rate, which will be driven mainly by the level of refi, which is driven by prevailing interest rates relative to the vintage of the underlying mortgages.
“canadian mortgages” How do they hedge out the “melting risk” on the IMBS (Igloo Mortage Backed Securities)?
NakedPuts Wrote: ------------------------------------------------------- > “canadian mortgages” > > How do they hedge out the “melting risk” on the > IMBS (Igloo Mortage Backed Securities)? That’s accounted for in the constant precipitation rate. Geezus can’t believe I’m gonna go ahead and hit “post messsage” - studying is getting to me!
thanks guys!
This is from the perspective of US mortgages…I dont’ know anything about canadian MBS. This should be a good exercise for you. Pull up the mortgage refinance index on bloomberg. BB ticker for the refi index is:{MBAVREFI Index } that will show you refi activity over time. Compare prepay activity on a universe of mortgages (all 30 year 6.00% FNMAs for example…bb ticker is {FNCL 6 n mtge cph cpr }) to the refi index. Look at CPRs over a period of fairly stable interest rates to get a feel for how prepays are affected by factors other than interest rate changes (ie housing turnover), then look at CPRs for the same universe for a period of falling interest rates to see the effect that interest rates have on prepays. You will see a tremendous difference. I’m not at my bloomberg right now but it wouldn’t surprise me if you pulled up a rates stable environment and you saw that prepay speeds on 30 year 6’s come in around 5 CPR, and that in a rates falling scenario those speeds could post well north of 55 CPR. Interest rates are a huge driver of refi activity. Another way you can look at this is to pull up a specific pool and run a yield table using the Blomberg Prepay Model. The base case under BPM shows how that model expects the prepays to come in under a “rates unchanged” scenario. The yield page will also give you up 100 to 300 and down 100 to 300 speeds. The BPM model is bloombergs best guess…you can use ADP, ESP, or PSA too. Ultimately this exercise will help point out that interest rates moves are a huge factor in prepays. If you want to see the actual vectors they are running you can pull it up after you’ve run the yield table using BPM. Type in {VBPM } and it will pull up the month by month speeds the model is using for each rate shock scenario. Another point to consider is that not all mortgage respond in the same degree to a given interst rate move. A 10 year borrower has much less incentive to refi than does a 30 year borrower. A few other points to consider, and that have helped to reduce the amount of refi activity one would normally expect under the interest rate scenario we’ve had over the last two years are: Credit standards have tightened LTVs are higher due to home price depreciation (many borrowers are underwater) Refi pipeline is no longer as efficient due to hundreds of mortgage companie failing Fewer exotic structures being underwritten to help people refi
>Credit standards have tightened LTVs are higher due to home price depreciation (many borrowers are underwater) Refi pipeline is no longer as efficient due to hundreds of mortgage companie failing Fewer exotic structures being underwritten to help people refi add refinancing burnout to that list since rates have been low vs. historical for some time.
people who posted on this thread, any other people who work with ABS: What are your job titles/responsiblities? I’m curious because I’m in risk mgmt and starting to work with ABS a little bit in terms of projecting losses on held securities and doing some valuation work on new issues we are looking at buying. I’d like to get more into this kind of work though the job market for it obviously isn’t that great right now. I’ve also worked with Student Loan TALF ABSes and I noticed they are extending TALF to commercial real estate so I might be doing some work there too. http://www.reuters.com/article/bondsNews/idUSN1737238320090817
juventurd, i work in the treasury dept of one of the big 5 canadian banks. main responsibilities are to conduct research on customer behavior to support the bank’s pricing and hedging models.