Week 40 Day 3: Risk vs. Uncertainty: Calculable vs. Unknowable

Risk: 'There is a 20% chance the stock drops 30%.' You can plan for that. Uncertainty: 'Something nobody has thought of will happen.' You cannot plan for that specifically, but you CAN build a portfolio resilient enough to survive it. Diversification is the defense against uncertainty: whatever happens, you own everything, so nothing can wipe you out.

The risk-uncertainty distinction (Frank Knight, 1921): (1) Risk: probabilities are known or estimable. Standard deviation, VaR, and Monte Carlo simulations all model risk. They assume the future will resemble a distribution based on the past. They work well for 'normal' market conditions. (2) Uncertainty (Knightian uncertainty): probabilities are unknown or unknowable. COVID-19 in 2020, 9/11 in 2001, the 2008 financial crisis (few models predicted the correlated collapse of housing, banks, and global credit simultaneously). These events existed outside the probability distributions that risk models used. Recent 'uncertainty' events and their market impacts: 9/11 (2001): S&P 500 dropped 12% in one week. Global financial crisis (2008): S&P 500 dropped 57% over 17 months. COVID-19 (2020): S&P 500 dropped 34% in 33 days. In each case, the event was largely 'unmodeled' -- it did not fit neatly into historical probability distributions. Defending against uncertainty: (a) Diversify across asset classes (stocks, bonds, cash, real estate, TIPS). Not all assets crash at once in every scenario. (b) Maintain an emergency fund (3-6 months cash). This prevents forced selling during uncertainty events. (c) Avoid leverage (leveraged losses in uncertainty events can be total). (d) Hold some TIPS or I-Bonds (protect against inflation uncertainty). (e) Keep your investment plan simple enough to survive scenarios you have not imagined.

Knight's (1921) distinction between risk (measurable) and uncertainty (unmeasurable) has profound implications for portfolio construction. Under risk, the optimal portfolio is the mean-variance efficient portfolio (Markowitz, 1952) -- calculable with known return distributions. Under uncertainty, the optimal strategy shifts toward robustness: performing adequately across a wide range of unknown scenarios rather than optimally under one assumed scenario. Taleb (2012) formalized this as 'antifragility' -- designing systems that benefit from disorder rather than merely surviving it. In portfolio terms, antifragile strategies include: (1) barbell allocations (very safe assets + very risky assets, avoiding the middle) that perform well in both normal and extreme environments, (2) systematic rebalancing (which profits from volatility by buying low and selling high mechanically), and (3) optionality (maintaining dry powder -- cash or short-term bonds -- to deploy during uncertainty-driven dislocations). Ellsberg (1961) showed that people are 'ambiguity averse' -- they prefer known risks to unknown uncertainties, even when the unknown uncertainty has the same or better expected value. This ambiguity aversion causes investors to overconcentrate in familiar investments (domestic stocks, large-cap companies, their employer's stock) and underinvest in unfamiliar ones (international stocks, small-caps, alternatives). The diversification prescription is therefore an uncertainty-management tool as well as a risk-management tool: by owning 'everything' (VTI + VXUS + BND + VTIP), the investor is robust to scenarios they cannot anticipate.