might be observed in plans that offer the full range of funds from a large
mutual fund company such as Fidelity or Vanguard (often called a “Mutual
Funds Window”). Such offerings are common in the 403(b) plans at univer-
sities and other nonprofits, but not in the corporate 401(k) plans in our
sample.
D. Alternative ExplanationsSo far, we have interpreted the positive correlation between the relative
number of equity options and the allocation to equities as supporting the
diversification heuristic. One concern, though, is that different equity funds
might serve different purposes. For example, adding a second growth fund
to a plan that already had a mix of equity funds should probably have lit-
tle effect on the overall asset allocation of the participants, but adding an
590 BENARTZI AND THALER
Table 16.7
The Relative Number of Equity-Type Investment Options and Asset Allocation:
A Regression Analysis
(Dependent Variable: The Percentage of Plan Assets Invested in Equities)Relative Indicator Log of
WLS Number Whether the the Plan
Regression of Equity Plan Offers Assets in
Model Intercept Options Company Stock Thousands Adjusted R^2
Panel A: No Industry Indicators (N=162)1 22.09 63.14 34.61 percent
(4.94) (9.28)
2 29.72 36.75 15.05 43.45 percent
(6.73) (4.49) (5.10)
3 10.57 36.77 14.78 1.40 44.16 percent
(0.89) (4.52) (5.03) (1.74)
Panel B: Including Industry Indicators Based on 2-Digit SIC Codes (N=142)
4 58.68 55.12 percent
(8.29)
5 43.90 12.93 58.91 percent
(5.39) (3.26)
6 47.07 9.09 4.13 61.79 percent
(5.93) (2.25) (2.96)
Notes: The initial sample consists of the June 1996 MMD sample of 401(k) plans. Eight
plans with less than four investment options were excluded, resulting in a sample of 162 plans.
When we include industry indicators, the sample is further reduced to 142 plans due to miss-
ing industry information. The table reports WLS regression estimates with plan assets as
weights (t-statistics are in parentheses).