Backcast/benchmark/valuate

From eg

Jump to: navigation, search

See risk as regret and gain/lose/reinvest for introduction. This is the living ontology pattern for practical use in decision-making and value creation, replacing single-scenario models (discounted cashflow, VaR and EVA). Quotes from "Seeing Tomorrow: weighing financial risk in everyday life", Ron S. Dembo and Andrew Freeman, outlining Dembo's risk as regret model, pp. 194-199, outlining the steps required to backcast/benchmark/valuate from a future fixed time horizon to decide today.

  1. "Rule 1: Know the value of your holdings today."
  2. "Rule 2: Pick an appropriate future time horizon." From which to verb:backcast.
  3. "Rule 3: Choose a wide range of scenarios to describe possible future events. Includes extremes and scenarios that contradict popular opinion. Include scenarios with negative outcomes that would cause regret. Assign a likelihood to each scenario."
  4. "Rule 4: Pick a benchmark."
  5. "Rule 5: Value your portfolio and benchmark at the horizon under each and every future scenario."
  6. "Rule 6: Compute the appropriate risk measure based on values obtained by applying Rule 5:"
    1. "Compute the difference in the horizon values between the portfolio and the benchmark under each scenario."
    2. "Regret is zero whenever this number is positive - we should feel happy about our portfolio in this case."
    3. "Regret is the absolute value of the difference whenever this value is negative."
    4. "Average regret is the probability-weighted sum of the regrets under each scenario."

"...regret is exactly what an insurer would have to pay you if he insured the downside on your portfolio relative to the benchmark. [An] insurer pays nothing if you outperform the benchmark and pays the absolute value of the difference if you underperform the benchmark. Thus the true value of this regret is the price you would have to pay for such insurance today...exactly the cost of eliminating all downside risk...since the benchmark can be arbitrary, regret can properly be used as a comparative measure, which is not the case for other risk measures such as VaR. For the value of regret you can eliminate risk entirely. Regret is therefore an appropriate measure for allocating your capital among businesses (or, if you are an individual, allocating your money between investments). We regard regret as a "perfect" measure of risk."

"The importance of our risk rules is that they separate the process of risk management into distinct and logical steps. The choice of scenarios is separate from the choice of methodology for valuation, which is in turn separate from the particular choice of risk measure." Thus backcast/benchmark/valuate. See also closely related patterns regret/reframe/reassure, assure/insure/ensure, focus/frame/filter, frame/match/mediate, gain/loss/capital and have/bet/get/go, value creation, intangibles and their measurement, etc.

"Combining regret with upside and then factoring in our appetite for risk using a measure such as lambda creates a powerful new way of approaching decisions. In many instances we will make better decisions because we will think more clearly about the array of outcomes that could happen in future. We can use scenarios to describe that array... by adopting the right benchmark we can change the way we think about many problems." Thus regret/reinvest/avert/insure; contrast with EVA and other single-scenario approaches.

"We will never eliminate regret. It is in the nature of things that bad, occasionally terrible, events happen that change our lives forever. Managers do make bad bets that ruin their careers. We might buy too much of an investment that collapses and ruins us. Plenty of us get our fingers burned in the housing market because we put too little weight on regret."

"But we can use regret against itself. If we know our potential regret, we can avoid situations in which we are exposed. Or we can minimize our regret by insuring against it. Where there are efficient markets, we can do this relatively cheaply. But sometimes we will pay whatever it takes to obtain peace of mind, to the extent that we might walk away from a deal entirely. It is this impulse that must dominate our thinking about risk."

Personal tools