Serving the Quantitative Finance Community

 
User avatar
Cuchulainn
Posts: 22934
Joined: July 16th, 2004, 7:38 am

Why Einstein Wouldn't Play Dice With Monte Carlo Risk Simulations?

November 25th, 2014, 4:53 pm

QuoteOriginally posted by: MBSADVISORSQuoteOriginally posted by: CuchulainnQuoteOriginally posted by: MBSADVISORS"Memory of a goldfish" I rest my case! Daveangel, You have been, and continue to be Hilarious!Re: Venn Diagram, Interesting,, What relationship values do you feel would have the most impact? A B C sets for example with D.. as the intersectionsD is the property and A,B,C, .., E, F, ... are stakeholders with an interest in D.D is not only an asset but also a liability, so does your model cater for the full lifecycle? There are large (deterministic) forces pressing down on D (and interacting with each other ...).Excellent this will help me a great deal in getting started, Thank you! I have some deadlines I am working on for clients, I will get to work on a first draft and post it once completed.BestBTW I have sent you a PM (left corner of screen).
 
User avatar
MBSADVISORS
Topic Author
Posts: 3
Joined: May 21st, 2012, 11:16 am

Why Einstein Wouldn't Play Dice With Monte Carlo Risk Simulations?

November 25th, 2014, 5:36 pm

Cuchulainn"BTW I have sent you a PM (left corner of screen)."Got it thank you
 
User avatar
Cuchulainn
Posts: 22934
Joined: July 16th, 2004, 7:38 am

Why Einstein Wouldn't Play Dice With Monte Carlo Risk Simulations?

November 26th, 2014, 8:44 am

QuoteOriginally posted by: MBSADVISORSCuchulainn"BTW I have sent you a PM (left corner of screen)."Got it thank youYou're welcome. Hope it was useful.Thinking out loud...Somewhere in the article is mentioned that property prices are not market-driven but is planned out by major stakeholder groups in a policy-driven approach.I reckon that these forces should be in a model? Deliberate policy-based land pricing makes things more predictable?Due to depreciation, the asset will eventually be written off but maybe that is outside the scope?How do you model volatility, going forward 20/30 years?
Last edited by Cuchulainn on November 25th, 2014, 11:00 pm, edited 1 time in total.
 
User avatar
MBSADVISORS
Topic Author
Posts: 3
Joined: May 21st, 2012, 11:16 am

Why Einstein Wouldn't Play Dice With Monte Carlo Risk Simulations?

November 26th, 2014, 3:30 pm

QuoteOriginally posted by: CuchulainnQuoteOriginally posted by: MBSADVISORSCuchulainnThinking out loud...Somewhere in the article is mentioned that property prices are not market-driven but is planned out by major stakeholder groups in a policy-driven approach.I reckon that these forces should be in a model? Deliberate policy-based land pricing makes things more predictable?Due to depreciation, the asset will eventually be written off but maybe that is outside the scope?How do you model volatility, going forward 20/30 years?A good question, which I will circle back to at a later time. First I believe we should start with a clean sheet of paper so we are all on the same page, absent of cluttered thoughts brought on by misconceptions about the amount of sub-prime loans generated and defaulted. I very much appreciate the members of this site bringing this issue to light, for I was not aware such disagreements were as commonplace as I originally thought. I am sure the media is largely to blame for this. Looking back through the numbers Subprime loans at their "peak" rose to just over 20% of all originations, of which 35% defaulted. Therefore Subprime mortgage defaults accounted for approximately 7% of the entire housing market at its peak. Please see attached stats and charts. Now look at The Cumulative Default rates for Fannie Mae, (Freddie Macs is virtually the same) at their peak (attached), their defaults surpassed those of subprime at 12% percent. Note the crisis began when the housing default rates were below 6%. Now lets view The Credit Characteristics of these same loan pools (provided). You will note between 2003 to 2005, as previously stated the underwriting guidelines were virtually the same, however defaults in 2003 were approximately 1%. Well within acceptable range. Leading to only one conclusion, a severe housing bubble would have occurred with or without the influence of subprime loans, for conventional loan defaults had risen to over 12%. All this being said, I wholeheartedly agree with other members, no doubt subprime mortgages made the crisis measurably worse than it would have been. If there are any lingering doubts, that the 30-year mortgage is "partly" to blame you need only look back to the last 3 major recessions when subprime mortgages were virtually nonexistent. Leading to the conclusion, something else was at hand, which is causing severe defaults approximately every 10 years. Now that all the stats are in plain view, it may be a good time to review the Simple Math from a fresh prospective here http://20-yearsimp.blogspot.com . Please view all Stats and charts attached.
Attachments
Total subprime orig at 9.31.48 AM.png.zip
(51.16 KiB) Downloaded 74 times
credit characteristics 2014-11-24 at 8.28.42 AM.png.zip
(196.48 KiB) Downloaded 58 times
1 fannie mae 2014-11-24 at 7.52.31 AM.png.zip
(120.06 KiB) Downloaded 69 times
Last edited by MBSADVISORS on November 26th, 2014, 11:00 pm, edited 1 time in total.
 
User avatar
MBSADVISORS
Topic Author
Posts: 3
Joined: May 21st, 2012, 11:16 am

Why Einstein Wouldn't Play Dice With Monte Carlo Risk Simulations?

November 29th, 2014, 4:54 pm

QuoteOriginally posted by: CuchulainnQuoteOriginally posted by: MBSADVISORSCuchulainnThinking out loud...Somewhere in the article is mentioned that property prices are not market-driven but is planned out by major stakeholder groups in a policy-driven approach.I reckon that these forces should be in a model? Deliberate policy-based land pricing makes things more predictable?Due to depreciation, the asset will eventually be written off but maybe that is outside the scope?How do you model volatility, going forward 20/30 years?This is a very interesting question. One that pretty much summarizes the reasoning behind all my posts. IF... the current Fed's MC based simulation models would have had better parameters which accounted for volatility they wouldn't have been (as Greenspan put it) "blindsided" in 2008 as the Feds were in 1979 and in 1989.The EG is a barometer which tracks potential volatility of the housing market based on HCTI index. Past stats outlined in my paper indicates the real-estate market has a large impact on the economy. The data from our research suggests HCTI ratios which approach 6:1 puts borrowers under financial distress leading to credit defaults which cause volatility in the housing market, decreases the circulation of disposable income, which slows down manufacturing, etc.... The EG model serves to put us on notice when the economy is on the wrong track, at which time we should look towards lowing HCTI levels, as fast as we can (what should we do to lower it? This is part of a much larger discussion). Unfortunately the EG is only part of the solution for it can only foresee the impending volatility, it will not solve it. My other posts deal with the major causes of volatility in the housing market and what we can do to solve it. For example The median income is being out paced by inflation 2:1 over ten years for the past four decades; the average borrowers mortgage is 50% of their net income. This leads to volatility due to the majority of borrower's will be under some sort of financial distress in several years. This is directly related to my post, The Obsolescence of The 30-year Mortgage. The 30-year models inability to retire principal fast enough to keep up with the effects of our modern day economy. The good news, this can be solved by introducing faster principal reduction models into the market. This will go a long way to negate the negative effects of 50% inflation over a 10 year period, by allowing borrowers to either sell or Refi their homes with 30% more equity, translating into 30% lower monthly payment to offset the effects of rapid inflation in our modern day economy. Our financial system has a lot of moving parts. Obviously there is no one silver bullet that will solve all the issues which made our financial system so volatile in the past. I have only mentioned a few of the major issues my research paper has uncovered. The EG model is the first step toward the ultimate goal of creating an algorithm which will track potential volatility far in advance; allowing us to put the necessary measures in place to negate such outcomes as we witnessed in 1979,1989, and 2008.
Last edited by MBSADVISORS on November 30th, 2014, 11:00 pm, edited 1 time in total.