Carl, a General Motors foreman, had saved regularly through his employer's payroll plan. His entire investment portfolio was in company shares. The company paid most of his family's income, too. When it entered bankruptcy in two thousand nine, Carl lost both his job and his portfolio, according to Burton Malkiel's account.
The family's wages and savings shared one vulnerability. Future earnings are part of your wealth, and an investment can become especially costly when it fails alongside your livelihood. Malkiel's answer is to reduce dependence on individual businesses, then judge new holdings by what they do to the risks already carried.
A failed product can damage one company. Owning many different businesses reduces that danger. A broad recession can damage them together. Buying more shares cannot remove that shared market risk. In the pricing model Malkiel explains, investors should expect compensation for bearing shared risk, with no extra reward for company risk they could cheaply diversify away. A concentrated bet may win, but its danger alone does not entitle you to higher returns.
The model relies on competition. If two diversified portfolios have the same market risk but different expected rewards, investors buy the more attractive one. Their buying raises prices and reduces the advantage. Keeping avoidable company risk brings no special claim to compensation.
Stocks have historically rewarded investors more than safer short-term holdings, with serious losses and long disappointing periods along the way. An expected return is an average of possible outcomes, weighted by their chances. It is never a promised payment. The useful question is whether an uncertain reward justifies a risk your household can afford.
What happens when you put them together
Malkiel's imaginary island makes the combination visible. It has an umbrella manufacturer and a resort. Rainy and sunny seasons are equally likely. In rain, the umbrella business earns fifty percent while the resort loses twenty-five percent. In sunshine, their fortunes reverse. Either business alone has an expected return of twelve and a half percent, with an uncomfortable loss in the wrong weather.
Put one dollar into each. Whatever the weather, one earns fifty cents and the other loses twenty-five. The combined gain is twenty-five cents on two dollars, or twelve and a half percent. Each business still fluctuates wildly. Together, in this deliberately tidy model, they preserve the expected return and remove the fluctuation.
Real businesses rarely offset each other perfectly. They don't have to. Returns that move together less closely can still reduce combined fluctuations. Diversification depends on economic differences. A carmaker and a tire supplier remain tied to automobile demand. The first varied additions to a concentrated portfolio usually help most; adding another holding to an already broad portfolio helps less.
The same logic allows a more volatile investment to make the whole steadier. In Malkiel's comparison covering nineteen seventy through twenty seventeen, developed foreign stocks fluctuated more than American stocks. Yet adding a modest foreign holding reduced the American portfolio's fluctuations and raised its historical average return. Their ups and downs did not coincide exactly. Adding too much foreign stock eventually raised risk again; the best historical mix was specific to that period.
Broad index funds, which track groups of investments, provide access to domestic and foreign stocks, bonds, and shares in property-owning companies. These investments respond to different economic forces. But those relationships change. Stock markets fell together during the financial crisis, and rising interest rates can hurt bonds and some stocks together. Having several categories on a statement does not establish that they will protect one another.
What the risk label leaves out
Volatility measures fluctuation, counting large gains as well as large losses. It cannot capture every danger a household faces. Another measure, beta, estimates sensitivity to market movements. Past beta can help describe relative fluctuation, but it has predicted returns poorly on its own. A number describing risk cannot reliably tell you what reward will follow.
That matters when a fund promises a better way to own the market. Conventional indexes give larger companies, measured by stock-market value, more weight. A factor strategy deliberately shifts those weights toward a characteristic associated with past returns. Favoring shares cheap relative to earnings, for example, may exploit neglected bargains, or expose you to companies in genuine distress. Both explanations can contribute; historical success does not settle which will matter next.
Nor does an attractive simulation show what an investor can keep. Academic strategies may require borrowing securities and trading frequently. Financing, trading, taxes, and fund fees can consume the apparent advantage. Popularity can raise the prices of favored holdings and reduce future rewards. Malkiel allows a place for inexpensive factor strategies, conditional on understanding their different risks and costs.
Even a portfolio split between stocks and bonds can get most of its fluctuation from stocks. Risk parity tries to spread contributions to risk more evenly, usually adding bonds. Some versions then borrow to raise expected returns. That borrowing magnifies losses, too. Rising rates make older bonds paying less interest less attractive, depressing their prices. The enlarged bond holding can suffer while changing relationships between assets undermine the assumed protection.
The test comes back to the household. In the same bad economy, what happens to your investments, your earnings, and your ability to wait? A broader portfolio helps only if its remaining risks are ones you can live through.
Three things to keep
Employer shares can put your wages and savings at risk together.
A volatile addition can steady a portfolio if its returns move differently enough.
Diversification leaves shared economic risks, uncertain rewards, and costs to judge.