There are two basic investment risk models, one based on projected cash flows over a long period of time, discounted at a variety of future interest rate scenarios, and one based on short term correlations of expected market values.? I call the first model the actuarial model, and the second the financial model (pejoratively, the Wall Street model).
Under ordinary conditions, the financial model looks better.? It asks, “Can we make money in the short run versus our capital costs?” The? actuarial model asks, “Can we assure that we will be solvent under a wide number of economic scenarios over the long run, some of which might be quite severe?”
During boom conditions, the financial model wins, while those following an actuarial model are branded fuddy-duddies.? During bust conditions, those following an actuarial model survive, while many following the financial model don’t.
There were many on Wall Street that claimed to be following a WOW “Worst Of the Worst” model.? I remember interviewing the chief risk officer of one of those firms in 2005 — Bear Stearns.? Talked a really good game.? To be fair, so did the risk manager of Goldman Sachs that year.? I assume most of the risk managers of Wall Street had their WOW models — after the crisis with LTCM, they had to look at the correlations on risk assets going to one in a crisis.
My guess is the WOW models were largely ignored, and the more common VAR models followed.? Perhaps Goldman And Morgan Stanley gave more weight to the worst outcomes, but hindsight is 20/20.? They might have survived in spite of themselves.
My point: you’ve got to survive in order to win.? Models that emphasize current profits at the expense of survivability get whacked during large busts.? Even if they survive, the hole that they must crawl out of is deep.
The economy is highly variable, and the financial economy as a derivative of it is even more so.? Companies that think long-term with respect to risk management tend to survive crises; they have limited their risks, and left returns on the table during the boom times.
Survival is a major part of the game.? Look at previously successful financial companies.? It doesn’t matter how well you did in the past if you are down 90, 95, 99% over the last two years.
As such, for those that invest in financial companies, evaluate their survivability.? How likely is it that they will get hit badly?? Are they overleveraged?? Do they need additional financing?
Actuarial models focus on the long run, and analyze survivability.? Why aren’t they used more frequently?? The actuarial models indicate a greater need for capital than VAR models.? More capital left in reserve means a lower return on equity, and a lower stock price in the short run.
High quality management teams for financials place more value on their long-run (actuarial) risk models.? They want to make money over the long term, if they can.? Those that focus on VAR will do better in the short run, until the next big bear market hits.? For value investors, stick with the quality players relying on long-term risk models.? Momentum players are free to play with the VAR users, but keep your stop orders ready.
You forgot that the most important aspect of any model is the Backstop. “What will I do when the s__t hits the fan?” Well, now we know what that plan is: taxpayer money.
The hole you speak of is perhaps deep, but the bankers are well above the opening and shareholders have prime seating just a few feet below the lip. My children and my children’s children are standing on the bottom watching the government and Wall Street dig the ground out beneath their feet.
paul wilmott addressed this back in nov
http://www.wilmott.com/blogs/paul/index.cfm/2008/11/17/Actuaries-Versus-Quants
“In the long run we are all dead” according to Keynes. While certainly true as regards individuals, it is less true as regards institutions. What has failed us are the actions of individuals who had no concern for the institutions that provided them with wages. Well, perhaps less so as regards Goldman Sachs and a few other firms that appear to be survivors. An important question for the financially adept is: Did you create a profitable enterprise; or, did you merely extract money from the market?
Actuarial models (at least for defined benefit pension evaluations)make guesses as to future events such as interest rates and mortality. At each evaluation the expectations are compared to what actually happened in the preceeding period and the guesses adjusted. Furthermore,a “cushion” is included to compensate for wrong guesses.
Why don’t economists and investors, indeed everyone, recognize that an Expected Value may never become an Actual Value; the future is always iffy. I read somewhere that primitive tribes maintain their populations so as to never need to use more than about 80% of their resources. And we think we’re smarter than they are?
Perthaps another way to say this is that Wall Street does best when it sticks to selling risk. That’s the “Wall Street” model.
On the other hand, Wall Street blows it when they start to think they are so good with risk that they can even buy it. That requires the “actuarial model”, and as we all have seen, Wall Street may have that model, but it doesn’t much seem to use it.
On the other hand, here’s the model of risk followed by (many) insurers in Florida, and signed off on by their actuaries.
500 execs run 50 companies that are priced to make money and grow aggressively in good years. In a WOW scenario, they are all out of business.
Four years later, those 500 execs are running 50 companies, priced the same way. Only the deck has been shuffled …
Don’t be too convinced that actuarial models are better than the Wall St. financial models. It comes down to the character of management in the end.
The problem with this is, when you really look under the cover of what the “actuarial models” propose to do, you see so many unjustified and frankly dangerous assumptions it makes your head spin. The stability of the insurance industry owes more to regulators who simply place vast conservative margins onto the reserves which the “actuarial models” would tend to produce.
The great great benefit of the “financial models” you speak of is that (when done correctly) they try to *imply* as many of the parameters as possible FROM THE MARKET — not from somebody’s fantasy world of what they “think” the equilibrium levels will be. I am speaking of things like levels of interest rates (use the forward curve), future equity returns (use the risk free rate), volatility (imply from option prices) future default rates (use the risk-neutral rates implied by market credit spreads), etc., etc. The complete meltdown of the financial system that we’re seeing today owes as much to the fact that people were willing to divorce themselves from the rigorous application of MARKET driven “financial modeling” and begin to make ad-hoc assumptions. If they had really followed the methods with rigor, you wouldn’t see all those fantasy marks on the banks books right now — they would be marked to MARKET (i.e., close to zero), and you would know who was insovlent and who wasn’t … and we’d be that much closer to turning the page on this crisis.
Just one last point — the “actuarial modeling” system is the system that has (A) brought the defined benefit pensions system to collapse globally, and (B) underlied all of the rating agency models of subprime ABS, CDO’s, etc., etc
Want to make your pension liabilities look manageable? No problem — just use the “actuarially justified” 8% equity rate of return to discount them! Want to make your crap assets AAA? No problem — just use historical default rates and correlations and assume everything “reverts to the mean” eventually.
No shortage of actuaries and their models at the following companies: AMBAC, MBIA, Countrywide (owned several P&C insurers they shuttled business to), AIG, HIG, etc.
Again, it really is NOT about whether the model is “financial” or “actuarial” it is about the character of the MANAGEMENT who uses, directs, and interprets the model.
Contra ‘been there, done that’ mean reversion is a pretty good principle, when applied with sufficient sophistication. Some common actuarial models have been pretty good guides
http://tinyurl.com/arg6j6
http://www.google.co.uk/search?hl=en&q=%22wilkie+model%22&btnG=Google+Search&meta=
David, this is a good article. Your classification of bull and bear markets and the use of actuarial and wallstreet models are good thoughts. I think the problem is not with bull and bear markets. It is with the predictability of bull and bear markets. The failure of Bear Sterns or Goldman Sachs models to account for the wild events that happened recently is because of the unpredictable nature of these events. We are talking about unprecedented events in financial markets. Even if you can predict such ‘black swans’ why would you provide for an event with a probability of less than 0.01%. Anyway, I found some interesting read in http://www.financialmodel.net as well.