THE MARKET TIMING ABILITY OF UK MUTUAL FUNDS Keith Cuthbertson*, Dirk Nitzsche*and Niall O’Sullivan** Abstract: We apply a recent nonparametric methodology to test the market timing skills of UK equity and balanced mutual funds. The methodology has a number of advantages over the widely used regression based tests of Treynor-Mazuy (1966) and Henriksson-Merton (1981). We find a relatively small number of funds (around 1%) demonstrate positive market timing ability at a 5% significance level while around 19% of funds exhibit negative timing and on average funds miss-time the market. However, controlling for publicly available information we find very little evidence of market timing ability based on private timing signals. In terms of investment styles, there are a small number of successful positive market timers amongst Equity Income and ‘All Company’ funds but not among either Small Stock funds or Balanced funds, although a few small stock funds are found to time a small stock index rather than a broad market index. Keywords : Mutual funds performance, market timing. JEL Classification: C14, G11 * Cass Business School, City University, London ** Department of Economics, University College Cork, Ireland Corresponding Author : Professor Keith Cuthbertson Cass Business School, City University London 106 Bunhill Row, London, EC1Y 8TZ. Tel. : +44-(0)-20-7040-5070 Fax : +44-(0)-20-7040-8881 E-mail : [email protected]We gratefully acknowledge the provision of mutual fund data by the Investment Research Partnership (www.tirp.co.uk). They acknowledge the use of the data provided under licence from Morningstar UK. We are grateful for financial support from the Irish Research Council for the Humanities and Social Sciences (IRCHSS). We thank an anonymous referee for helpful comments and suggestions. Main programmes use Gauss TM .
Transcript
THE MARKET TIMING ABILITY OF UK MUTUAL FUNDS
Keith Cuthbertson*, Dirk Nitzsche*and Niall O’Sullivan** Abstract: We apply a recent nonparametric methodology to test the market timing skills of UK equity and balanced mutual funds. The methodology has a number of advantages over the widely used regression based tests of Treynor-Mazuy (1966) and Henriksson-Merton (1981). We find a relatively small number of funds (around 1%) demonstrate positive market timing ability at a 5% significance level while around 19% of funds exhibit negative timing and on average funds miss-time the market. However, controlling for publicly available information we find very little evidence of market timing ability based on private timing signals. In terms of investment styles, there are a small number of successful positive market timers amongst Equity Income and ‘All Company’ funds but not among either Small Stock funds or Balanced funds, although a few small stock funds are found to time a small stock index rather than a broad market index. Keywords : Mutual funds performance, market timing. JEL Classification: C14, G11 * Cass Business School, City University, London ** Department of Economics, University College Cork, Ireland Corresponding Author : Professor Keith Cuthbertson
Cass Business School, City University London 106 Bunhill Row, London, EC1Y 8TZ.
We gratefully acknowledge the provision of mutual fund data by the Investment Research Partnership (www.tirp.co.uk). They acknowledge the use of the data provided under licence from Morningstar UK. We are grateful for financial support from the Irish Research Council for the Humanities and Social Sciences (IRCHSS). We thank an anonymous referee for helpful comments and suggestions. Main programmes use Gauss
There is now a well established literature evaluating mutual fund performance, much of
which seeks to determine whether funds truly add value to investors through superior or
abnormal returns. This question of performance is important, not least because investors
are allocating their (scarce) resources to such funds often as saving instruments for the
future. The question of performance therefore has policy implications, given a predicted
future pensions gap (Turner 2004, OECD 2003). Recent UK studies have examined fund
performance in relation to security selection skill (Cuthbertson et al 2008, Keswani and
Stolin 2005, Fletcher and Forbes 2002, Quigley and Sinquefield 2000) and have also
applied parametric factor regressions to examine market timing (Byrne, Fletcher and
Ntozi 2006, Fletcher 1995).
Attempts at market timing involve either tactical asset allocation, the use of
derivatives or rebalancing the fund’s risky equity holdings to increase (decrease) the fund
market beta in response to an expected bull (bear) market.1 In this paper we apply the
nonparametric procedure of Jiang (2003) to examine the market timing performance of
UK domestic Equity funds and Balanced funds. We employ a large survivorship-bias free
data set of over 800 (non-tracker, non second-unit) funds.
The nonparametric procedure has several advantages over the parametric
(regression) approach. First, it measures the quality of a fund manager’s timing
information rather than the aggressiveness of his response - whereas the widely used
regression based methods of Treynor-Mazuy (TM) (1966) and Henriksson-Merton (HM)
(1981) do not separate these two elements2. The quality of timing information is of more
interest to the investor as he can control the aggressiveness of his position himself,
simply by adjusting his holdings of risky/non-risky assets. In addition, unlike the TM and
HM tests, the nonparametric approach examines timing over multiple frequencies (daily,
monthly, quarterly etc) and does not assume that the timing frequency is uniform or even
known, across funds. Many existing timing tests may have low power when actual fund
timing frequencies differ from data sampling frequencies (Goetzmann et al 2000, Bollen
and Busse 2001).
1 UK mutual funds are restricted in their use of derivative securities since the assets of the fund must be able to fully cover any liabilities that are created when employing derivative contracts. In practice this prevents the fund from achieving any real gearing and ensures that the fund is able to meet its liabilities if called upon to do so. 2 The Treynor-Mazuy and Henriksson-Merton tests of market timing are well documented and widely applied
in the literature. See, for example, Fletcher and Schadt (1996). To conserve space we do not outline these methods here.
2
Several further difficulties arise with the TM and HM tests. Breen at al (1986)
show that the HM regression may exhibit heteroscedasticity and that ignoring this renders
the HM test poor both in terms of size and power. 3 An additional difficulty with the TM
and HM tests concerns their inability to decompose overall fund abnormal performance
into its market timing and security selection components, (Admati et al 1986, Grinblatt
and Titman 1989). Many studies point to a negative correlation between the market
timing and selectivity measures of performance (Jagannathan and Korajczyk 1986,
Coggin et al 1993, Goetzmann et al 2000, Jiang 2003). For example, Jiang (2003) reports
simulation results showing a negative correlation between the two performance measures
in the TM and HM models, even where none exists, whereas the correlation between the
nonparametric timing measure and the security selection measure in the regression
models is very small (indistinguishable from zero for larger sample sizes). Jagannathan
and Korajczyk (1986) suggest that a spurious negative correlation may arise due to the
nonlinear pay-off structure of options and option-like securities in fund portfolios - holding
a call option on the market yields a high pay-off in a rising market but in a steady or falling
market the premium payment lowers return and appears as poor security selection4.
Conditional market timing regression based tests, control for timing ability that
may be attributable to public information and provide a test of private timing skill by funds.
Ferson and Schadt (1996) specify the time-varying portfolio beta to be a function of a set
of predictive public information variables – this gives rise to a conditional market timing
model. Conditional timing may also be measured by the correlation between a private
timing signal and market returns (Becker et al 1999, Byrne et al 2006). In this paper, we
also apply tests of conditional market timing but in a nonparametric framework.
The bulk of the US empirical evidence on market timing demonstrates no market
timing or perverse negative market timing (Wermers 20005, Ferson and Schadt 1996,
Becker at al 1999, Goetzmann et al 2000, Jiang 2003) - although conditioning on public
information is shown to improve the model specification (Ferson and Warther 1996,
Ferson and Schadt 1996, Becker at al 1999). Mamaysky et al (2008) use the Kalman
filter to model time varying betas (and alphas). With dynamic estimates the authors
explore which trading strategies are associated with outperformance. The findings
indicate that superior and inferior returns are linked to attempts at market timing rather
than stock selection, though overall there is little evidence that investors earn superior
3 For further discussion on the power of standard regression based tests of abnormal performance see
Kothari and Warner (2001). 4 The returns on the common stock of highly geared firms may create a similar effect.
5 Wermers (2000) examines market timing using holdings data and controls for size, book-to-market and
momentum effects.
3
returns. Bollen and Busse (2005) examine persistence in market timing and find evidence
of short term persistence but only when using daily data.
In the relatively sparse literature evaluating market timing skills among UK mutual
funds, Byrne, Fletcher and Ntozi (2006) is an important recent contribution, using the
parametric (regression) approach6. The authors evaluate the conditional timing
performance of UK unit trusts using a method which incorporates benchmark investment
by funds and fund risk aversion to deviations from benchmarks. Byrne et al (2006) find
that benchmark investment is important to UK funds and funds are highly risk averse to
deviations from the benchmark. The authors find no evidence of positive conditional
market timing among individual funds or portfolios of funds. Fletcher (1995) applies both
the Chen and Stockum (1986) test (similar to TM) and the HM test. Evaluating 101 unit
trusts between 1980 and 1989, Fletcher reports the cross sectional average timing
measures to be negative and strongly significant for both models of market timing and for
alternative market benchmark indices. Leger (1997) evaluates UK equity investment
trusts between 1974 and 1993 and finds similar results - negative and statistically
significant market timing.
The central contribution of this paper is to apply a set of non-parametric tests of
market timing to a large database of UK equity and balanced mutual funds domiciled in
the UK and also to UK funds domiciled in offshore locations. This, provides
complementary evidence to existing UK studies, which are based on parametric
regression approaches. The paper proceeds as follows: Section 2 describes the
nonparametric testing methodology. In section 3 we describe the UK data set, empirical
results are reported in section 4 and section 5 concludes.
2. Nonparametric Test of Market Timing
Because of the difficulties discussed above with regression based tests of market timing,
Jiang (2003) proposes a non-parametric test (applied to US mutual funds), which we
outline briefly below. The market model is:
(1) i,t+1 i i,t m,t+1 i,t+1r =α +β r +ε
where i,t+1r is the excess return on fund i, m,t+1r is the relevant benchmark excess return,
iα is a security selectivity measure (assumed to be independent of market timing) and
the fund beta, i,tβ , is assumed to vary with the fund manager’s market timing information
at time t. The fund’s timing skill is determined by the ability to correctly predict market
6 Byrne et al (2006) apply the conditional timing methodology of Becker et al (1999).
4
movements. Let m̂,t+1 m,t+1 tr =E(r |I )be the manager’s forecast for the next period’s market
return based on the information set tI , and define the parameter v as:
The next step is to link the manager’s forecast of the market return with their response in
adjusting i,tβ in (1). For any triplet of market return observations 1 2 3m,t m,t m,t{r ,r ,r } sampled
from any three time periods (not necessarily in consecutive order) with 1 2 3m,t m,t m,t{r <r <r } ,
an informed market timer will maintain a higher exposure to the market over the
2 3m,t m,t[r ,r ] range than in the1 2m,t m,t[r ,r ] range. Nonparametric beta estimates for both time
ranges are 1 2 1 2 1t i,t i,t m,t m,tβ =(r - r )/(r - r ) and
2 3 2 3 2t i,t i,t m,t m,tβ =(r - r )/(r - r ) . Here beta embodies
both the precision of the market return forecast and the aggressiveness of the manager’s
response where the latter is affected by risk aversion. Grinblatt and Titman (1989) show
that for a fund i with non-increasing absolute risk aversion and independent timing and
selectivity information ˆ
t
m,t+1
β> 0
r
yielding a convex fund return/market return relationship
(4) 3 2 2 1
3 2 2 1
i,t i,t i,t i,t
m,t m,t m,t m,t
r - r r - r>
r - r r - r
which allows (3) to be written as 2 1 2 1t t m,t +1 m,t +1v =2 x Pr(β >β |r >r ) -1. A sample statistic of
a fund’s timing ability may be constructed as:
(5) ˆ 3 2 2 1
3 2 2 1m,t m,t m,t1 2 3
-1
i,t i,t i,t i,t
n
m,t m,t m,t m,tr <r <r
r - r r - rnθ = sign >
3 r - r r - r
5
where sign () = (1, -1, 0) for positive, negative and zero market timing respectively. ˆ nθ is
the average sign across all triplets taken from n observations. ˆ nθ can be shown to be n-
consistent and asymptotically normal (Abrevaya and Jiang 2003, Serfling 1980) with
variance:
(6) ˆˆˆ
1 2 3n
1 2 3 1 2 1 3
2-1n
2t t t nθ
t =1 t <t ,t ¹t ,t ¹t
n9σ = h(z ,z ,z ) - θ
2n
where
(7) 3 2 2 1
1 2 3 1 2 3
3 2 2 1
i,t i,t i,t i,t
t t t m,t m,t m,t
m,t m,t m,t m,t
r r r rh(z ,z ,z ) sign | r r r
r r r r
Under the null hypothesis of no market timing ˆˆ ˆ
nn θ
z = n.θ σ is asymptotically N(0,1)7.
As discussed, one difficulty in examining a fund’s market timing skill is
distinguishing the quality of the manager’s forecast of the future market return from the
aggressiveness of response in changing the fund beta. The TM and HM market timing
measures do not separate out these two elements. The nonparametric measure on the
other hand simply measures how often a manager correctly forecasts a market
movement and acts on it - irrespective of how aggressively they act on it. The sign
function in (5) assigns a value of 1(-1) if the argument is positive (negative) regardless of
the size of the argument.
As previously discussed, a further advantage of the nonparametric measure is
that it is robust in testing for timing skill among managers whose timing frequency may
differ from the frequency of the sample data and/or whose timing frequency may not be
uniform or even known. The timing statistic in (5) investigates timing over all triplets of
fund returns rather than just consecutive observations and consequently uses more
information than parametric tests8. Finally, the HM regression approach suffers size and
power distortion under heteroscedasticty but the asymptotic distribution of the
7 Note, the calculation in (6) includes permutations
1 2 3 2 1 3 3 1 2t t t t t t t t th(z ,z ,z ),h(z ,z ,z ),h(z ,z ,z ) , i.e. the
same three market return observations drawn in different combinations. However, the sign in (7) is equal in all
three cases since it is conditional on 1 2 3m,t m,t m,tr <r <r . That is, irrespective of the order in which the market
return observations are drawn they are first sorted in ascending order and there can only be one such sorting. 8 See Ferson and Khang (2002) and Ferson, Kisgen and Henry (2006) for a discussion of the problem of
interim trading bias.
6
nonparametric timing measure in (5) is unaffected by heteroscedasticity in fund returns
(Abreveya and Jiang 2005).9
The nonparametric test embodies some relatively mild restrictions on behaviour.
The test requires tβ be a non-decreasing function of m̂,t+1r . Grinblatt and Titman (1989)
demonstrate that this requires non-increasing absolute risk aversion. This is less
restrictive than that of the TM and HM measures which require specific linear and binary
response functions, respectively. For example, the linear response function embodied in
the TM measure is consistent with the manager maximising a Constant Absolute Risk
Aversion (CARA) preference function (Admati et al, 1986). However, such an assumption
is questionable if there is non-linearity in the payment to fund managers in respect of
benchmark evaluation (Admati and Pfleiderer, 1997), option compensation (Carpenter,
2000) and a non-linear performance-flow responses by investors (Chevalier and Ellison,
1997).
Conditional Market Timing: Public versus Private Information
The nonparametric test can be applied as a conditional statistic after allowing for market
timing skill attributable to public information. This conditional measure involves first
calculating both sets of residuals from regressions of the mutual fund returns and market
returns on the lagged public information variables. Clearly, these residuals represent the
variation in the fund and market returns not explained by the public information. Denoting
the pair-wise fund and market regression residuals as i,tr and m,tr respectively, the
procedure described above in (5) may then be applied to the residuals to yield a
conditional timing measure
(8) 3 2 2 1
3 2 2 1m,t m,t m,t1 2 3
-1
i,t i,t i,t i,t
n
m,t m,t m,t m,tr <r <r
r - r r - rnθ = sign >
3 r - r r - r
Note, ˆ nθ in (5) and nθ in (8) can clearly be of different magnitudes but may also be of
different sign. For example, ˆnθ >0 but nθ 0 may indicate a successful market timing
manager whose skill is attributable entirely to public information.
9 Abrevaya and Jiang (2005) show that for a wide range of possible “curvatures”, the non-parametric z-
statistic is superior to alternatives, while the “size” and use of the asymptotic standard deviation is accurate (even when heteroscedasticity is present) for sample sizes as low as 25.
7
3. Data
Our mutual fund data set contains monthly returns on 842 (actively managed) UK equity
Unit Trusts and Open Ended Investment Companies and 174 UK Balanced funds. ‘UK
Equity’ funds must have at least 80% of the fund capital invested in UK domestic equity –
though typically this figure is closer to 95% for most funds. This data set represents
almost the entire set of UK equity and balanced funds which existed at any point during
the period January 1988 – December 200210. Equity funds are classified by investment
objectives: Equity Income funds (162), ‘All Company’ funds (553) and Small Stock funds
(127). Investment objectives are declared by funds but are certified initially and monitored
monthly by the Investment Management Association (IMA). Equity Income funds aim to
have a dividend yield in excess of 110% of the yield of the FT All Share Index. All
Company funds select from a broad universe of UK domestic equity. Small Stock funds
have at least 80% of the fund invested in UK equities which form the bottom 10% of all
equities by market capitalisation. Similarly, our set of UK Balanced funds invest in UK
assets only. These funds are classified by Standard and Poors (S&P) into a number of
categories as follows: “UK Neutral” - funds with a portfolio invested in the asset classes of
equity, fixed interest and money market securities with an equity content always in the
range of 30% - 70% of the portfolio. “UK Flexible” - funds with a portfolio that may invest
in any proportion in any asset class. “UK Dynamic” – funds with a portfolio invested in the
asset classes of equity, fixed interest and money market with an equity content usually
above 70% of the portfolio. “UK Defensive” - funds with a portfolio invested in the asset
classes of equity, fixed interest and money market but typically with no more than 30% in
equities. We follow the usual convention in using net returns (bid-price to bid-price, with
gross income reinvested) so returns are those that accrue to the investor after the annual
charge (and trading costs of the fund) but before any front or back-end loads.
Fund ‘second units’ have been excluded from the analysis. These arise for the
most part when a single fund is sold under different pricing structures to different groups
of investors such as retail and institutional or when the same fund is sold under slightly
different pricing structures through life assurance companies etc. Second units do not
represent separate independent portfolios and hence we exclude them. Standard and
Poors assign a unique identification number to all funds: the S&P ID. The S&P ID is
helpful in tracing a fund’s history through name changes and mergers with other funds
etc. Name changes can be problematic as it can lead to the inadvertent inclusion of the
same fund twice. In this study, the S&P ID code of every fund was examined and name
changes and mergers were identified in order to avoid this error. Examination of the S&P
ID codes reveals that many funds whose history cease before the end of the sample
10
Data Source: Morningstar UK.
8
period, ie ‘nonsurvivors’, were in fact taken over by other funds and did not necessarily
close due to poor performance.
The data set includes both surviving funds (775) and nonsurviving funds (241) in
order to control for survivorship bias. In addition, funds are also categorized by the
location of operation - onshore funds (836) are domiciled in the UK while offshore funds
(180) are domiciled in locations such as Dublin, Luxembourg, Isle of Man, Channel
Islands and some other European locations, although all funds invest in UK assets. We
later examine onshore and offshore funds separately for differences in performance.
The market benchmark is the FT All Share Index of total returns (i.e. including
reinvested dividends) taken from Datastream. We also evaluate funds’ timing
performance against style specific benchmarks. The style specific benchmark used here
for Equity Income funds is the Morgan Stanley Capital International (MSCI) UK value
index. The MSCI value index is available through the MSCI website. In constructing this
index MSCI adopts a two-dimensional style segmentation (into value and growth indices)
in which securities are assigned to each index by z-scores based on different attributes.
Value attributes include the dividend yield, book-to-market ratio and 12 month forward
earnings to price ratio. As the value index is comprised in large part of high dividend yield
securities we adopt this index as a style specific benchmark for Equity Income funds.
The style specific index employed for Small Stock funds is the Hoare-Govett Smaller
Companies Index which measures the performance of the lowest 10% by market
capitalisation of the main UK equity market. The Hoare-Govett Smaller Companies index
is produced by Elroy Dimson and Paul Marsh of London Business School and published
by ABN-AMRO. As All Company funds select from a broad universe of UK domestic
equity we simply use the FT All Share index as the market benchmark here.
In our conditional market timing tests we use conditioning variables as follows: the
UK one-month T-Bill return, the dividend yield on the FT All Share index and a measure
of the term spread (i.e. difference between the UK 30 year Gilt and the one-month T-Bill
yields). We also consider a broader range of public information variables than
considered in previous studies, including leading indicators that are closely watched by
financial market professionals. Here we include: a Bond-Equity return ratio, the month-on-
month percentage change in the OECD composite leading indicator for the UK, the CBI
Industrial Trends Survey order book volume balance and month-on-month percentage
changes in UK retail sales, industrial production and the retail price index (ex-mortgage
payments). Our bond index is the Citigroup UK Government Bond index obtained from
Morningstar UK. This fixed-income index measures the total return for bonds with a
9
remaining maturity of at least one year. Each bond has a minimum size criterion to avoid
liquidity biases. Total returns are market-capitalization-weighted. All other conditioning
variables are taken from Datastream (whose original sources include OECD, Office for
National Statistics and the Confederation of British Industry).
4. Empirical Results
The unconditional market timing results are presented in Table 1. Row 1 displays the
market timing test statistic, ˆˆ ˆ
nn θ
z = n.θ σ at various points in the cross-section of
performance ranging from the lowest to highest. The z-statistic is asymptotically N(0,1)
under the null hypothesis of no market timing. Row 2 displays the market timing
coefficient, ˆnθ , corresponding to the fund in row 1.11 In our discussion of results,
‘statistical significant’ refers to a 5% significance level (one-tail test) unless stated
otherwise. – check this is one sided test ? kc ##
Table 1 here
From the z-statistic in row 1, it is evident that there are only a small number of skilled
market timers. From the full set of results, the “top” 8 (18) funds demonstrate statistically
significant positive market timing ability at a 5% (10%) significance level (one-tail test) –
around 1% (2%) of the sample of funds12. The cross-sectional average test statistic is z =
-0.752 while an equally weighted portfolio of all funds yields a significant test statistic of z
= -1.9808. More specifically, 78% of funds demonstrate negative market timing while 19%
are statistically significant negative market timers. Overall, the nonparametric test fails to
find evidence of positive timing ability among more than a ‘handful’ of UK mutual funds
while on average, funds are found to miss-time the market. As a robustness test we also
perform the nonparametric test using the “FT 100 return” as the market benchmark. Table
1, row 5, shows the sorted market timing test statistic. The findings are qualitatively
similar to those using the FT All Share returns where only a small number of funds (two)
are found to successfully time the market at 5% significance and on average funds miss-
time the market.
11
To improve statistical reliability, results are reported for funds with a minimum of 36 observations which leaves 794 funds in the analysis. 12
When discussing the proportion (or total number) of funds that have a statistically significant value for z , then strictly speaking we are in a multiple testing framework so the significance level for the overall proportion
of significant funds will be different from the 5% significance level for each fund taken individually (because of compound type-I errors) – see Barras, Scaillet and Wermers (2009).
10
For comparison, Table 1 (row 3 and row 4) also reports the t-statistics of the market
timing coefficients of the TM and HM tests (for the funds sorted as in row 1)13.
Interestingly, 7 of the top 8 funds which are found to be statistically significant positive
market timers using the nonparametric test are also found to be successful market timers
using the TM (HM) procedure. However, from the full set of results the regression tests
suggest stronger evidence of market timing, since for the TM and HM models 27 and 17
funds respectively are found to have statistically significant positive timing skill, whereas
the result for the non-parametric test is 8 funds. The greater prevalence of positive market
timing found by the TM and HM measures may arise because these methods also
incorporate the aggressiveness of the market timing response, unlike the nonparametric
test.
To mitigate survivorship bias we include nonsurviving funds in the analysis. Of the
794 funds examined, 211 are nonsurvivors. In Table 1, the row denoted ‘Survival’
indicates whether the sorted funds were survivors or nonsurvivors where, 1 = survivor, 0
= nonsurvivor. Nonsurviving funds are approximately equally represented among positive
and negative market timers. However, in the tails of the cross-section of performance
distribution the results differ: none of the funds which demonstrate statistically significant
positive timing ability is a nonsurvivor and of the top 20 sorted funds only two are
nonsurvivors. In the left tail of the performance distribution, 43 of the 148 funds (29%)
which show significant negative market timing are nonsurvivors.
Our (unconditional) market timing results for UK mutual funds are broadly in line
with those of Jiang (2003) for the US who reports that between 2% and 5% of funds
possess statistically significant positive timing skill (depending on the alternative market
indices used) and also reports that the average US fund displays negative timing ability.
As a simple examination of whether market timing ability is related to the age of
the fund, the final row of Table 1 reports the number of (monthly) observations for each of
the funds. It is evident that better performing market timers are generally shorter-lived
funds – this is broadly consistent with the Berk and Green (2004) model where superior
(market timing) skill may lead to both higher average returns in the short-run and higher
inflows (chasing winners), with the latter eventually leading to diseconomies of scale and
hence less successful market timing skills of older (more established) funds14. However,
when we group funds by age-bracket, the average market timing test statistic among
13
The TM and HM, t-statistics are based on Newey-West heteroscedasticity and autocorrelation adjusted standard errors. 14
Barras, Scaillet and Wermers (2009) posit a similar reason to explain why older US funds have lower
alphas than younger funds.
11
funds with between 3 and 5 years of observations is z = -0.493 while among funds of 10
years or more is z = -0.917 - although both figures are statistically insignificant. Jiang
(2003) also reports negative and insignificant market timing (on average) among similar
age categories of US funds. This highlights the importance of looking a individual fund
performance.
Market Timing Performance by Investment Style and Location
We now examine possible differences in timing skill between funds with different
investment objectives, i.e. Equity Income, All Company, Small Stock and Balanced funds.
We first evaluate timing ability relative to the FT All Share index and then apply style
specific benchmarks where appropriate. Using the broad market benchmark there is
some potential for spurious timing inferences across fund investment styles. One difficulty
is the assumed independence between security selection and market timing information.
A manager’s information in both these areas may be correlated and consequently
selectivity and market timing inferences may be difficult to ‘disentangle’ (Admati et al
1986, Grinblatt and Titman 1989). For example, it has been argued that small stock
funds may exhibit spurious timing against a market benchmark comprised of large stocks,
as small stocks may have option-like characteristics, (Jagannathan and Korajczyk, 1986).
Alternatively, it may be argued that ‘All Company’ (general equity) funds select from the
broadest universe of stocks which make up the benchmark market portfolio, again
creating a possible overlap between selectivity and timing decisions.
Table 2 here
Notwithstanding these caveats, Table 2 reports the market timing results by
investment style in each panel. Row 1 presents the nonparametric z-statistics for funds
evaluated against the FT All Share index. There is some evidence of positive timing
ability for both Equity Income funds and All Company funds in the extreme right tails of
the distribution, while no Small Stock funds or Balanced funds exhibit statistically
significant positive market timing. However, forming portfolios of funds by investment
style yields nonparametric test statistics of –1.162, -2.265, -2.468 and –0.803 for Equity
Income, All Company, Small Stock and Balanced funds, respectively. This shows that in
the case of Equity Income funds and All Company funds while there are a small number
of top market timers, as groups these classes of funds perform poorly on average
(significantly negative in the case of All Company funds). In contrast, there are no skilled
market timers among the group of Balanced funds but as a group these funds perform
12
the least badly. Against the broad market index Small Stock funds produce no skilled
market timers and also perform (significantly) poorly on average as a group.
Figure 1 provides a graphical illustration of these findings where the distributions
of Equity Income and All Company funds show a small number of skilled timers in the
extreme right tails but both are positively skewed, while the Small Stock fund distribution
is generally to the left of the others. In the Balanced funds group both tails of the
distribution reveal less extreme good and bad performers15.
Figure 1 here
We also investigate whether the fund classifications result in timing a specific style
benchmark. In Table 2, Panel A, the row denoted “z (MSCI Value)” shows the
nonparametric test statistic for Equity Income funds evaluated against a value stock
index. (See section 3 for discussion on the style specific indices). Only one fund shows
significant timing ability but overall the results are qualitatively similar to those from the
broader market-wide index. From Panel C, row 5, denoted “z(Hoare Govett)”, however,
Small Stock funds are found to be more adept at timing a small capitalisation benchmark
rather than our market benchmark. The cross-sectional distribution for z(Hoare Govett)
lies further to the right of the distribution for z(FT All) . The values for z(Hoare Govett)
indicate that 5 Small Stock funds successfully time the small-cap index and there is
considerably less statistically significant negative market timing. As a group, small stock
funds on average positively time the small stock index, although not significantly so. As
All Company and Balanced funds select from a broad universe of stocks we do not
evaluate these funds against a style specific equity index. 16
Table 3 here
The performance of onshore and offshore funds is evaluated separately in Table
3. Here, we examine whether geographical proximity gives rise to informational
asymmetries and performance differences (Otten and Bams 2007, Brennan and Cao
1997, Coval and Moskowitz 1999). Panel A presents results for the 672 onshore UK
funds while Panel B reports results for the 122 offshore funds. A small number of both
15
In Panel D of Table 2, the row denoted “Style” further sub-classifies balanced funds where 1 = Defensive funds, 2 = Neutral funds, 3 = Flexible funds and 4 = Dynamic funds. (See section 3 for definitions). From the results no sub-class of balanced funds emerges as notably good or bad market timers. For example, each sub-class is well represented among both the top and bottom 20 balanced funds. 16
As the investment styles of our funds are already clearly defined and monitored by the Investment
Management Association we select a single representative style benchmark for all funds within a style rather than constructing an individual benchmark for each fund. Results in Byrne et al (2006) when using the “Sharpe style index” give qualitatively similar results to those when using the market index in the parametric regression approach.
13
onshore and offshore funds (around 1% and 2% respectively) exhibit statistically
significant positive market timing when using the nonparametric z-statistic, while among
onshore funds a higher proportion of funds exhibit significant negative market timing
(20%) compared to 13% of offshore funds. Previous UK studies have found differences in
selectivity skill between onshore and offshore funds (Cuthbertson et al 2008). Here,
however, there is no evidence that the more geographically distant offshore funds
underperform in terms of market timing. This may be because there is less or no
informational asymmetry when predicting ‘macro’ level market movements compared to
the ‘micro’ level security selection required for generating positive alpha.
Conditional Market Timing
As discussed in section 2, conditional market timing tests can be used to control for
timing skill attributable to publicly available information and hence can determine whether
evidence of successful market timing is attributable to private timing signals on the part of
funds. Investors are likely to be more interested in identifying funds with private market
timing performance in excess of that achievable by public information signals alone.
Conditioning public information variables to forecast market return predictability, usually
include lagged values of (i) a short term government bond yield, (ii) the market dividend
yield, (ii) the spread in the term structure, (iv) the credit spread as well as seasonal
effects such as a January dummy, (Ferson and Schadt 1996, Byrne et al 2006). In this
study we use the UK one-month TBill return, the dividend yield on the FT All Share index
and a measure of the term spread (i.e. the difference between the UK 30 year Gilt and
the one-month TBill yields). We also consider a broader range of public information
variables than typically considered in previous studies including leading indicators that
are closely watched by financial market professionals. We include the Bond-Equity return
ratio calculated as the ratio of the total return on a broad index of UK government bonds
to equity market total returns. A bond-equity return ratio measures the relative
attractiveness of these two asset classes and may help predict institutional capital flows
and market movements (Clare, Thomas and Wickens 1994). We then consider two
further leading economic indicators: First, the OECD composite leading indicator for the
UK which is designed to provide early signals of turning points in economic activity. The
components of this index cover a range of short term economic indicators as well as
consumer and business survey based opinion. We use the month-on-month percentage
change in the index. Second, the Confederation of British Industry (CBI) Industrial Trends
Survey. Here, we use the CBI order book volume balance. This balance is a weighted
measure of firms’ responses to whether orders are expanding or contracting and hence
provides a leading signal of macro-economic activity and possibly of market movements.
14
Finally, we consider month-on-month percentage changes in UK retail sales, industrial
production and the retail price index (ex-mortgage payments) as having possible market
predictability. We control for a possible January effect using a dummy variable.
Table 4 here
We present the results of predictive regressions in Table 4. The predictive
variables are specified with a lag except for the January dummy which is
contemporaneous. The first column of results shows OLS estimates and their t-statistics
from bivariate regressions of the market return on the predictive variables taken
separately. (t-statistics are White (1980) adjusted as appropriate). The second column
shows the results of a multivariate regression. In both sets of results the market dividend
yield and the bond-equity return ratio are statistically significant while in the multivariate
regression the TBill return, the term spread and the OECD composite leading indicator
also show some evidence of explanatory power. Our findings relating to the dividend
yield are similar to those of Byrne et al (2006) while the findings in relation to the
significance of the term spread differ between the two studies – possibly due to the
different overall set of variables in the multivariate regressions.
Table 5 here
In our conditional tests of market timing we use the larger set of public information
instruments found to be significant (at the 5% level) in the multivariate regression in Table
4 - thus we control as much as possible for the public information contribution to timing
ability. Results are presented in Table 5 where row 2 shows the nonparametric
conditional z-statistics corresponding to the fund’s unconditional z-statistics in row 1. Only
one of the funds found to successfully time the market using the unconditional z-statistic
(the 7th) shows evidence of timing skill (at a 5% significance level) after controlling for
public information. (The overlap is 5 funds if we use a 10% significance level). With only
one exception, our results overall indicate no evidence of skillful market timing,
attributable to private information by funds.
5. Conclusion
For the first time on UK data, we apply non-parametric tests to assess the market timing
performance of individual UK mutual funds (as well as portfolios of funds) using a large
survivorship free data base of around 800 (non-tracker, non second-unit) funds. The
non-parametric method should be more informative than standard regression based tests
15
as it is based only on the quality of the manager’s timing signals rather than the
aggressiveness of his response – it is the former which is of greater interest to investors.
On the basis of our non-parametric tests we find that a relatively small number
(around 1%) of UK mutual funds possess significant positive market timing skill, while
around 19% are shown to miss-time the market. Evidence of positive market timing
ability using the non-parametric approach is found to be less frequent than for regression
based approaches – this may be, because the latter confounds both the quality of the
signal and the aggressiveness of the manager’s response to timing signals. On
conditional market timing, we find that after controlling for publicly available information,
there is very little evidence of successful market timing based on private information by
individual UK funds.
Overall, the nonparametric approach implies little evidence of successful market
timing among UK mutual funds. This result generally supports the findings of past UK
studies (Byrne et al 2006, Fletcher 1995, Leger 1997). One possible explanation for our
results lies in the impact of cashflow on fund behavior (Bollen and Busse 2001, Edelen
1999, Ferson and Warther 1996). In a rising market, funds may experience increased
investor cash inflows, a relatively high (short term) cash position and hence lower overall
exposure to the market and lower returns. Conversely, in a falling market redemptions
may increase, causing the fund to reduce its cash position leading to higher market
exposure. A further possible explanation may lie in the interdependency between timing
the level and volatility of the market where even if a fund manager expects a high market
return he may not increase the portfolio’s market exposure without also considering
market volatility (Chen and Liang 2006, Busse 1999). Finally, conditional tests of market
timing implicitly test a joint hypothesis: that the chosen public information variables have
market predictability and these same instruments are used by fund managers. However,
we use a large set of predictor variables so it seems unlikely that our finding of very few
successful conditional market timers amongst UK mutual funds is due to exclusion of
valid predictor variables.
16
References Abrevaya, J. and W. Jiang (2003), Pairwise slope statistics for testing curvature, working paper, University of Chicago Graduate School of Business. Abrevaya, J. and W. Jiang (2005), A nonparametric approach to measuring and testing curvature, Journal of Business and Economic Statistics, January, Vol. 23, No.1, pp 1-19. Admati, A., S. Bhattacharya and P. Pfleiderer (1986), On timing and selectivity, Journal of Finance, Vol. 41, No. 3, pp. 715-730. Admati, A. and P. Pfleiderer (1997), Does it all add up? Benchmarks and the compensation of active portfolio managers, Journal of Business, Vol. 70, No. 3, pp. 323-350. Barras, L., O. Scaillet and R. Wermers (2009), False Discoveries in Mutual Fund Performance: Measuring Luck in Estimated Alphas, forthcoming Journal of Finance. Becker, C., W. Ferson, D. Myers and M. Schill (1999), Conditional market timing with benchmark investors, Journal of Financial Economics, Vol. 52, No. 1, pp. 119-148. Berk, J.B. and R. C. Green, 2004, Mutual Fund Flows and Performance in Rational Markets, Journal of Political Economy, 112, 1269-95. Bollen, N. and J. Busse (2001), On the timing ability of mutual fund managers, Journal of Finance, Vol. 56, No. 3, pp. 1075-1094. Bollen, N. and J. Busse (2005), Short term persistence in mutual fund performance, Review of Financial Studies, Vol. 18, No. 2, pp. 569-597. Breen, W., R. Jagannathan and A. Ofer (1986), Correcting for heteroscedasticity in tests for market timing ability, Journal of Business, Vol. 59, No. 4, pp. 585-598.
Brennan, M. and H. Cao (1997). International portfolio investment flows, Journal of Finance, Vol. 52, Issue 5, pp.1851-80.
Busse, J. (1999). Volatility Timing in Mutual Funds: Evidence From Daily Returns, The Review of Financial Studies, Winter 1999, Vol. 12, No. 5, pp. 1009 – 1041. Byrne, A., Fletcher, J. and P. Ntozi (2006). An exploration of the conditional timing performance of UK unit trusts, Journal of Business Finance and Accounting, 33 (5) & (6), pp 816-838. Carpenter, J. (2000), Does option compensation increase managerial risk appetite, Journal of Finance, Vol. 55, No. 5, pp. 2311-2331. Chen, C. and S. Stockum (1986), Selectivity, market timing and random behaviour of mutual funds: a generalized model, Journal of Financial Research, Spring, pp. 87-96. Chen, Y. and B. Liang (2006). Do Market Timing Hedge Funds Time the Market. Discussion Paper, available at SSRN. Chevalier, J. and G. Ellison (1997), Risk taking by mutual funds as a response to incentives, Journal of Political Economy, Vol. 105, No. 6, pp. 1167-1200.
17
Clare, A., S. Thomas and M. Wickens (1994), Is the gilt-equity yield ratio useful for predicting UK stock returns?, The Economic Journal, Vol. 104, No. 423, pp. 303-315. Coggin, D., F. Fabozzi and S. Rahman (1993), The investment performance of US equity pension fund managers: an empirical investigation, Journal of Finance, Vol. 48, No. 3, pp. 1039-1055. Coval, Joshua D., and Tobias J.Moskowitz, 1999, Home Bias at Home: Local Equity Preference in Domestic Portfolios, Journal of Finance 54, 2045-2074. Cuthbertson, K., D. Nitzsche and N. O’ Sullivan (2008), Mutual fund performance: skill or luck?, Journal of Empirical Finance, Vol. 15, Issue 4, pp. 613-634. Edelen, R.M. (1999), Investor flows and the assessed performance of open-end mutual funds, Journal of Financial Economics, Vol. 53, No. 3, pp. 439-466. Ferson, W. and K. Khang (2002), Conditional performance measurement using portfolio weights: evidence for pension funds, Journal of Financial Economics, Vol. 65, Issue 2, pp. 249-282. Ferson, W., Kisgen, D. and T. Henry (2006), Evaluating government bond performance with stochastic discount factors, Review of Financial Studies, Vol. 19, No. 2, pp. 423-455. Ferson, W. and R. Schadt (1996), Measuring fund strategy and performance in changing economic conditions, Journal of Finance, Vol. 51, No. 2, pp. 425-462. Ferson, W. and V. Warther (1996), Evaluating fund performance in a dynamic market, Financial Analysts Journal, Vol. 52, No. 6, pp. 20-28. Fletcher, J. (1995), An examination of the selectivity and market timing performance of UK unit trusts, Journal of Business Finance and Accounting, Vol. 22, No. 1, pp. 143-156. Fletcher, J. and D. Forbes (2002), An exploration of the persistence of UK unit trust performance, Journal of Empirical Finance, Vol. 9, No. 5, pp. 475-493. Goetzmann, W., J. Ingersoll Jr. and Z. Ivkovich (2000), Monthly measurement of daily timers, Journal of Financial and Quantitative Analysis, Vol. 35, No. 3, pp. 257-290. Grinblatt, M. and S. Titman (1989), Portfolio performance evaluation: old issues and new insights, Review of Financial Studies, Vol. 2, No. 3, pp. 393-421. Henriksson, R. and R. Merton (1981), On market timing and investment performance II: statistical procedures for evaluating forecasting skills, Journal of Business, Vol. 54, No. 4, pp. 513-33. Jagannathan, R. and R. Korajczyk (1986), Assessing the market timing performance of managed portfolios, Journal of Business, Vol. 59, No. 2, pp. 217-235. Jiang, W. (2003), A nonparametric test of market timing, Journal of Empirical Finance, Vol. 10, No. 4, pp. 399-425. Keswani, A. and D. Stolin (2005), Mutual Fund Performance Persistence and Competition: A Cross-Sector Analysis, Journal of Financial Research, Vol. 29(3), September, pp 349-366. Kothari, S. and J. Warner (2001), Evaluating mutual fund performance, Journal of Finance, Vol. 56, No. 5, pp. 1985-2010.
18
Leger, L. (1997), UK investment trusts: performance, timing and selectivity, Applied Economics Letters, Vol. 4, No. 2, pp. 207-210. Mamaysky, H., M. Spiegel and H. Zhang (2008), Estimating the dynamics of mutual fund alphas and betas, Review of Financial Studies, January, Vol. 21, No. 1, pp. 233-264. Newey, W. and K. West (1987), A simple positve semi-definite heteroscedasticity and autocorrelation consistent covariance matrix, Econometrica, Vol. 55, pp. 703-708. OECD, 2003, Monitoring the Future Social Implication of Today’s Pension Policies, OECD, Paris, unpublished.
Otten, R. and D. Bams (2007). The Performance of Local versus foreign mutual fund managers, European Financial Management, Vol. 13, No. 4, pp. 702-720.
Quigley, G, and R. Sinquefield (2000), Performance of UK Equity Unit Trusts, Journal of Asset Management, Vol. 1, No. 1, pp. 72-92. Serfling, R. (1980). Approximation theorems of mathematical statistics, Wiley, New York. Treynor, J. and K. Mazuy (1966), Can mutual funds outguess the market?, Harvard Business Review, Vol. 44, No. 4, pp. 131-136. Turner, Adair, 2004, Pensions : Challenges and Choices : The First Report of the Pensions Commission, The Pensions Commission, The Stationary Office, London. Wermers, R. (2000), Mutual fund performance: an empirical decomposition into stock-picking talent, style, transactions costs and expenses, Journal of Finance, Vol. 55, No. 4, pp. 1655-1695. White, H. (1980), A heteroscedasticity-consistent covariance estimator and a direct test for heteroscedasticity, Econometrica, Vol. 48, pp. 817-838.
19
Table 1: Mutual Fund Market Timing Performance – Unconditional Tests Table 1 presents the results of the unconditional market timing tests for selected points in the cross-sectional distribution. Row 1 reports the nonparametric test statistic,
ˆˆ ˆ
nn θ
z = n.θ σ which is asymptotically distributed as N(0,1) under the null hypothesis of no market timing. Funds are sorted from best to worst based on this statistic. Row 2
reports ˆnθ , the market timing coefficient, of the funds in row 1 while row 3 and row 4 show the corresponding (Newey-West adjusted) t-statistics of the TM and HM timing
coefficients respectively. Row 5 presents the sorted nonparametric test statistic, z, using the FT100, rather than the FT All Share index, as the market benchmark. Row 6 describes the investment objective of the funds in row 1 where, 1 = Equity Income fund, 2 = All Company fund, 3 = Small Stock fund, 4 = Balanced Fund. Row 7 indicates whether the fund is a survivor or a nonsurvivor: 1 = survivor, 0 = nonsurvivor. Row 8 describes the fund location: 1 = onshore, 0 = offshore. Row 9 displays the number of monthly observations. Results relate to the period 1988M1:2002M12 and are restricted to funds with a minimum of 36 observations, leaving 794 funds in the analysis.
Unconditional Market Timing Results
min 5.min min5% min10% min40% max30% max10% max5% max3% 20max 15max 12max 10max 7max 5.max 3.max 2.max max
Table 2: Mutual Fund Market Timing Performance – By Fund Type Table 2 presents the results of the unconditional market timing test statistic by investment style for selected points in the cross-sectional distribution. In each panel,
Row 1 reports the nonparametric test statistic, n
ˆn θˆ ˆz = n.θ σ , sorted from best to worst. Row 2 reports ˆnθ , the market timing coefficient of the funds in row 1 while
row 3 and row 4 show the corresponding (Newey-West adjusted) t-statistics of the TM and HM timing coefficients respectively. In Panel A, row 5 presents the test statistic, z, sorted from best to worst using the MSCI Value index as the market benchmark. In Panel C, row 5 presents the sorted test statistic, z, using the Hoare Govett Small Cap (HGSC) index as the market benchmark. In all panels, rows denoted ‘survival’ indicate whether the fund is a survivor or nonsurvivor: 1 = survivor, 0 = nonsurvivor. Rows denoted ‘Location’ indicate fund location: 1 = onshore, 0 = offshore. The final row in each panel displays the number of monthly fund observations. In Panel D on balanced funds, the row denoted ‘style’ indicates the sub-classes of balanced funds: 1 = Defensive, 2 = Neutral, 3 = Flexible, 4 = Dynamic. All results relate to the period 1988M1:2002M12 and are restricted to funds with a minimum of 36 observations leaving 143 Equity Income funds, 423 ‘All Company’ funds, 109 Small Stock funds and 119 Balanced Funds.
Unconditional Market Timing – By Fund Type
Panel A : Equity Income
min 5.min min5% min10% min20% min40% max30% max20% max10% 10max 7max 5.max 3.max 2.max max
Figure 1: Distributions of the Unconditional Market Timing Test Statistic – By Fund Type Figure 1 shows histograms of the cross-section of unconditional market timing test statistics, z, by fund type as indicated. The figures are based on 143 Equity Income, 423 All Company, 109 Small Stock and 119 Balanced funds with at least 36 monthly observations.
23
Table 3: Mutual Fund Market Timing Performance – By Fund Location Table 3 presents the results of the unconditional market timing test statistic by fund location for selected points in the cross-sectional distribution. Row 1 reports the
nonparametric test statistic, ˆˆ ˆ
nn θ
z = n.θ σ , sorted from best to worst. Row 2 reports ˆnθ , the market timing coefficient of the funds in row 1 while row 3 and row 4
show the corresponding (Newey-West adjusted) t-statistics of the TM and HM timing coefficients respectively. Row 5 indicates whether the fund is a survivor or nonsurvivor: 1 = survivor, 0 = nonsurvivor. Row 6 describes the investment objective of the sorted funds: 1 = Equity Income fund, 2 = All Company fund, 3 = Small Stock fund, 4 = Balanced Fund. Row 7 displays the number of fund observations. Results relate to the period 1988M1:2002M12 for funds with a minimum of 36 observations leaving 672 onshore and 122 offshore funds.
Unconditional Market Timing – By Investment Location
Panel A : Onshore UK Funds
min 5.min min5% min10% min20% min40% max20% max10% 20max 15max 10max 5.max 3.max 2.max max
Test Stat, z -3.522 -3.007 -2.389 -2.064 -1.617 -1.080 0.026 0.417 1.066 1.239 1.458 1.812 2.801 2.861 3.868
Table 4: Predictive Regressions of UK Equity Market Returns Table 4 presents the results of predictive regressions where the dependent variable is the FT All Share returns. The first column of results shows OLS coefficient estimates from bivariate regressions of the market returns on each of the predictive regressors taken separately. The second column shows the results of a multivariate regression. t-statistics are in parentheses – White (1980) adjusted as appropriate. R1 is the UK one-month T-Bill return , DY is the dividend yield on the FT All Share index, TERM is the term spread calculated as the difference between the UK 30 year Gilt and T-Bill returns, BEYR is the Bond-Equity return ratio calculated as the ratio of returns on a broad index of UK government bonds to equity market returns, OECD is the month-on-month change in the composite leading indicator for the UK produced by the OECD, CBIORD is the Confederation of British Industry Order Book Volume (Balance), RS is the month-on-month percentage change in UK retail sales, IP is the month-on-month percentage changes in UK industrial production, RPIX is the monthly percentage change in the UK Retail Price Index (excluding mortgage interest) and Jan is a January dummy variable. The predictive variables are specified with a lag except in the case of the January dummy which is specified contemporaneously. Results relate to the period 1988M1 – 2002M12.
Bivariate
Regressions Multivariate Regression
Predictive Variables
R1,t-1 0.117 (0.966)
-1.630 (-3.201)
DYt-1 0.840 (2.473)
4.448 (3.850)
TERMt-1 -0.013 (-0.084)
-1.849 (-3.172)
BEYRt-1 -0.146 (-2.074)
-0.202 (-2.871)
OECDt-1 0.738 (0.902)
1.805 (2.408)
CBIORDt-1 -0.002 (0.099)
0.032 (1.340)
RSt-1 -0.396 (-1.019)
-0.367 (-1.008)
IPt-1 0.103 (0.291)
0.127 (0.373)
RPIXt-1 0.081 (0.428)
-0.099 (-0.371)
Jant 1.237 (1.054)
0.435 (0.369)
R2
n/a 0.13
26
Table 5: Mutual Fund Market Timing Performance – Conditional Tests Table 5 presents the results for the conditional market timing tests at various points in the cross-sectional distribution. For ease of comparison, row 1 repeats the unconditional test statistics reported in Table 1. Row 2 reports the nonparametric test statistics of the conditional market timing test corresponding to the funds as
sorted in row 1. The conditioning variables are the UK one-month T-Bill return, the dividend yield on the FT All Share index, the term spread, the Bond-Equity return ratio and the OECD composite leading indicator for the UK. Row 3 describes the investment objective of the funds in row 1 where, 1 = Equity Income fund, 2 = All Company fund, 3 = Small Stock fund, 4 = Balanced Fund. Row 4 indicates whether the fund is a survivor or a nonsurvivor: 1 = survivor, 0 = nonsurvivor. Row 5 describes the fund location: 1 = onshore, 0 = offshore. Row 6 displays the number of monthly observations. Results relate to the period 1988M1:2002M12 and are restricted to funds with a minimum of 36 observations, leaving 794 funds in the analysis.
Conditional Market Timing
min 5.min min5% min10% min40% max30% max10% max5% max3% 20max 15max 12max 10max 7max 5.max 3.max 2.max max
