64  Risk and the Cost of Capital 
chance it will be worth $28 billion, for a net value of $13 billion. But there is 
a 20 percent chance it will fail to receive regulatory approval and be worth 
zero, leading to a loss of $15 billion. The expected value is $7 billion net of 
investment.10 Failure will bankrupt the company, because the cash flow from 
the company’s existing plants would be insufficient to cover its existing debt 
plus the debt on the failed plant. In this case, the economics of the nuclear 
plant spill over onto the value of the rest of the company. Failure would wipe 
out all the equity of the company, not just the $15 billion invested in the plant.
The implication is that a company should not take on a risk that will put 
the rest of the company in danger. In other words, don’t do anything that has 
large negative spillover effects on the rest of the company. This caveat would 
be enough to guide managers in the earlier example of deciding whether to 
go ahead with project A. If a $2,000 loss would endanger the company as a 
whole, management should forgo the project, despite its 60 percent likelihood 
of success. But by the same token, companies should not avoid risks that don’t 
threaten their ability to operate normally.
Executives making decisions for their companies should think about the 
company’s risk profile, not their own.11 After all, that’s the job of corpora-
tions; they are designed to take risks and overcome the natural loss aversion 
of individuals. The earliest corporations were the British and Dutch East India 
shipping companies. With those, if a ship sank, all shareholders would lose 
a tolerable amount instead of having one ship owner lose his entire fortune.
Professors Daniel Kahneman and Amos Tversky have demonstrated that 
most people place greater weight on the potential economic losses from their 
decisions than on the potential equivalent gains. In a McKinsey survey of 1,500 
global executives across many industries,12 we presented the executives with 
the following scenario: You are considering making a $10 million investment 
that has some chance of returning, in present value, $40 million over three 
years, with some chance of losing the entire investment in the first year. What 
is the highest loss you would tolerate and still proceed with the investment?
A risk-neutral executive would be willing to accept a 75 percent chance of 
loss and a 25 percent chance of gain. One-quarter of $40 million is $10 million, 
which is the initial investment, so a 25 percent chance of gain creates an ex-
pected risk-neutral value of zero. But most survey respondents demonstrated 
extreme loss aversion; they were willing to accept only a 19 percent chance of 
loss to make this investment, nowhere near the risk-neutral answer of 75 per-
cent. In fact, only 9 percent of respondents were willing to accept a 40 percent 
10 The calculation is ($13 billion × 80%) + (−$15 billion × 20%).
11 “The remainder of this section is adapted from D. Lovallo, T. Koller, R. Uhlaner, and D. Kahneman, 
“Your Company Is Too Risk-Averse,” Harvard Business Review (March–April 2020), hbr.org.
12 T. Koller, D. Lovallo, and Z. Williams, “Overcoming a Bias against Risk,” McKinsey & Company 
(August 2012), https://www.mckinsey.com/business-functions/strategy-and-corporate-finance/our-
insights/overcoming-a-bias-against-risk. 

Decide How Much Cash Flow Risk to Take On  65
or greater chance of loss. Informally, we’ve asked groups of executives the 
same question at even lower levels of investment and found similar results. 
Our findings echo those from Professor Ralph O. Swalm, going back to 1966.13
This phenomenon has serious consequences for hierarchical organizations. 
Executives are just as loss-averse when the bets are small as they are when the 
gambles are large, even though small gambles do not raise the same issues of 
survival or ruin that provide a rationale for aversion to large risks. What’s more, 
small gambles offer opportunities for the risk-reducing effects of aggregation.
To overcome loss aversion and make better investment decisions, individ-
uals and organizations must learn to frame choices in the context of the entire 
company’s success, not the individual project’s performance. In practice, this 
means looking at projects as a portfolio by aggregating them, rather than fo-
cusing on the risk of individual projects.
One technology company successfully used a portfolio approach to as-
sess its projects. First, executives estimated the expected return of each project 
proposal (measured as expected present value divided by investment) and the 
risks associated with each (measured as the standard deviation of projected 
returns). Executives then built portfolios of projects and identified combina-
tions that would deliver the best balance between return and risk. When they 
viewed portfolios of projects in the aggregate (Exhibit 4.4), executives could 
13 These results build on a 1966 Harvard Business Review article, “Utility Theory: Insights into Risk Taking,” 
by Ralph O. Swalm. He studied executives with varying levels of spending authority and found that risk-
preference profiles were very similar for executives at different levels of the organization.
EXHIBIT 4.4  Aggregating Projects Reduces Risk While Achieving High Expected Returns
Projects
Return, 
ratio of present value to 
investment
Risk, 
standard deviation of expected 
return, %
Expected net 
present value, 
$ million
4.5
15
8,100
A
15.4
64
200
12.4
B
104
500
7.5
C
66
50
4.7
E
150
200
4.4
F
52
500
3.7
G
37
30
3.7
H
29
400
2.7
I
58
900
2.6
J
31
400
2.5
K
150
300
2.3
L
20
220
1.9
M
18
520
1.5
N
20
300
1.1
O
13
850
0.9
P
5
2,000
0.3
Q
5
850
4.7
D
22
5
Portfolio of selected 
projects A–Q