Saturday, April 1, 2006, 07:22 PM
Working on yet another claims experience analysis I cam to the conclusion that I am rather obssessive. Though it should be fairly routine as there's not that much to finding the value of IBNR's, I'm always looking for things to pick on. I got rather cranky that I didn't have enough data about the fund to run a 'complete' analysis, the history which was so complex I was uncertain how to put it together.
On the train mulling this, I figured that my approach to claims experience analysis is identical as my apporach to men. I always think about the one experience I'm analysising and worrying about past events, unsure of how to report my results to get a good forecast on the future. Sometimes I would get so engrossed in a trivial thing like calculating the average amount of GSC claims I would lose sight of the big picture.
But at the end of the day at work, feeling irritated that I haven't gotten through it, I have to shut down my computer - and I realise that I need to do that too with men.
Like work, you have to make sure that obssessing over relationships doesn't overtake your life. And though my work is something that keeps me going day by day, I must always maintain enough perspective to step back and take in the big picture.
After all, in the end, it's just another claims experience analysis. And no matter how thrilling it is, it's just a part of all that is out there - there are valuations and pricing to be done!
And though the last time I totally made the wrong assumptions and set my parameters incorrectly, I must remember; it's just another claims experience.
[ 20 comments ] ( 246 views ) | [ 0 trackbacks ] | permalink |
( 3 / 215 )Saturday, April 1, 2006, 07:15 PM
We've talked about many psuedo-random things, but what point does one
go from dating to relationship phase? Seeing someone special, the marginal utility of each date increases - but where is teh pareto optimal efficient point on the dating frontier? Plus, don't forget, you must weigh up the opportunity cost of other relationships the utility of expected emotional benefit and the large variances attached to such.
And at this point in time, do we use implicit or explicit differentiation? Assume the relationship to be true and prove by inducation or proclaim the relationship as an axiom amd build equations from there?
For me, the problem is sensitivity testing. How do I set the parameters just right? Usually with my exceptionally blunt methods, everything goes out of whack and I need to reset the program. But unfortuantely, I can't just reload the spreadsheet in relationships. It's like an exceptionally bad claims experience in one year that drives up my premium prices forever more.
In fact, I read a book on dating called "Mars & Venus on a Date" (research for the column, you understand...) which claims that there are 5 stages in the dating process.
1. Attraction
2. Uncertainty
3. Exclusivity
4. Intimacy
5. Engagement
At first I got very excited at this as I could potentially model this as a Markov Procoess, run a Monte-Carlo simulation and see the long run probablities of which state I'll end up in. But alas - the next state of the relationship is dependent on all previous states with varying transition rates - failing the Markov property.
So here I am wondering if the next time I open my mouth, how will it affect the rate of change to our relationship.
*sigh* But as they say, the higher the risk, the higher the return... and I need that return due to inflation of pressures the older I grow.
[ 18 comments ] ( 107 views ) | [ 0 trackbacks ] | permalink |
( 3 / 213 )Back Next
