1/ I ran a course on applied financial #modeling for #credit relationship managers last week.
In the day and age of #models, do #analysts still need to understand how models are put together and numbers projected.
Credit systems give us numbers. Do we need more?
#banking #risk
In the day and age of #models, do #analysts still need to understand how models are put together and numbers projected.
Credit systems give us numbers. Do we need more?
#banking #risk
2/ To translate model figures into useful insights, an analyst must understand what drives a business and how those drivers would be impacted by future outlook.
A model is only as good as the input we plug in. Without relevant insight good models also produce garbage
A model is only as good as the input we plug in. Without relevant insight good models also produce garbage
3/ For instance, how would remote learning impact margins at school systems? Would enrollment numbers go up or down? How about non-core expenses?
When a senior #credit committee member asks a question or challenges an assumption, you can't hide behind the #model.
When a senior #credit committee member asks a question or challenges an assumption, you can't hide behind the #model.
4/ Putting together a credit risk model comes with its own challenges. Here are a few things you can do to ease your suffering.
First give models the commitment and patience they deserve. Don't rush into building them.
Measure what matters. Identify metrics you want to track
First give models the commitment and patience they deserve. Don't rush into building them.
Measure what matters. Identify metrics you want to track
5/ Model funding /debt schedule/ each line separately. Don't mix or blend.
Model fixed asset schedule separately. Including provisions for future capital expenditure and salvage value sales.
Don't use cash bank balances or equity as a balancing item. Model them separately
Model fixed asset schedule separately. Including provisions for future capital expenditure and salvage value sales.
Don't use cash bank balances or equity as a balancing item. Model them separately
6/ You need segregated models for the three items above because you need to run cross checks for model robustness. Without the segregated model tracing why your model is out is a nightmare
Even worse. Not knowing your model is out because everything balances.
Even worse. Not knowing your model is out because everything balances.
7/ Quick cross checks.
a) Tally net assets against retained earning or equity.
b) Tally change in cash on balance sheet with change in cash on statement of cash flows.
c) Compare historic common sized benchmarks against projected benchmarks. Super efficiency is an error.
a) Tally net assets against retained earning or equity.
b) Tally change in cash on balance sheet with change in cash on statement of cash flows.
c) Compare historic common sized benchmarks against projected benchmarks. Super efficiency is an error.
8/ But all of this goes to naught if you haven't done your homework on business drivers, economic outlook, assumptions, market structure and competition.
Validate your assumptions by taking your model for a dry run.
Validate your assumptions by taking your model for a dry run.
9/ Based on real life cases from Founder Puzzles - on the zen of financial modeling for founders
https://twitter.com/rebootdude/status/1287048108988411906?s=20
• • •
Missing some Tweet in this thread? You can try to
force a refresh
