More pot luck than skill to pick­ing a top fund

Even if you se­lect and in­vest in a unit trust fund that per­forms con­sis­tently, you will most likely end up earn­ing the av­er­age re­turn over time, an in­de­pen­dent re­searcher has found. Laura du Preez re­ports

Weekend Argus (Saturday Edition) - - PERSONALFINANCE - Daniel Wes­sels

It is bet­ter to in­vest in a lo­cal eq­uity unit trust fund that pro­duces av­er­age re­turns con­sis­tently than in a fund that at times per­for ms bet­ter and at times worse than av­er­age, an in­de­pen­dent re­searcher says.

Ideally, you should in­vest in funds that achieve above-av­er­age re­turns con­sis­tently. The dif­fi­culty is that you can­not ac­cu­rately pre­dict which fund will do so, Daniel Wes­sels, a part­ner at Cape Town fi­nan­cial ad­vi­sory com­pany Martin Ek­steen Jor­daan Wes­sels, says.

His study of how con­sis­tently broad-based do­mes­tic eq­uity unit trust funds per­for m from one year to the next found that, on av­er­age, funds have a 32-per­cent chance of main­tain­ing their per­for­mance, a 34-per­cent chance of do­ing bet­ter and a 34-per­cent chance of per­for ming worse. (Broad-based do­mes­tic eq­uity funds in­vest across all sec­tors of the JSE.)

How­ever, a few of the funds did pro­duce ex­cep­tion­ally con­sis­tent per­for­mance over five-or seven-year pe­ri­ods, Wes­sels says.

He looked at the rel­a­tive per­for mance of all the funds in do­mes­tic eq­uity gen­eral, growth and value sub-cat­e­gories for each 12-month pe­riod from March 31, 2002 to March 31, 2009.

In par­tic­u­lar, he con­sid­ered how the funds fared when they were ranked on the ba­sis of their per­for­mance and how, when the rank­ings were di­vided into four quar­tiles (quar­ters), the funds moved be­tween th­ese quar­tiles from one year to the next.

Based on his anal­y­sis, Wes­sels con­cluded that the rel­a­tive per­for­mance and rank­ing of lo­cal eq­uity funds exhibit ran­dom­ness year-on-year.

But he found that funds in the top quar­tile (those among the top 25 per­cent of funds ranked by per­for­mance) on av­er­age were more likely to main­tain their po­si­tion among the top 25 per­cent than were funds that fea­tured else­where in the rank­ings.

Wes­sels also found that if funds were awarded points de­pend­ing on in which of the four quar­tiles their per­for­mance ranked, with higher points for bet­ter per­for­mance, it would demon­strate which funds per­for med well con­sis­tently. The PlexCrown Fund Rat­ings, which Per­sonal Fi­nance pub­lishes on our unit trust prices page, take into ac­count the con­sis­tency of a fund’s per­for­mance. The rat­ing com­bines four dif­fer­ent mea­sures of per­for mance that take into ac­count the risk the fund took to achieve its re­turn over pe­ri­ods of up to five years.


When Wes­sels looked at the per­for­mance of the eq­uity funds rel­a­tive to the two lo­cal eq­uity mar­ket in­dices – the JSE All Share in­dex (Alsi) and the Share­holder Weighted in­dex (Swix) – he found that their per­for mance typ­i­fied the ac­tive ver­sus pas­sive in­vest­ing de­bate.

Lo­cal in­dices, par­tic­u­larly the Alsi, con­tain a size­able por­tion of shares in the min­ing and re­sources sec­tor, whereas lo­cal ac­tively man­aged eq­uity funds tend to be un­der­weight in this sec­tor and in in­di­vid­ual shares that con­sti­tute a large por­tion of the mar­ket, such as BHP Bil­li­ton and An­glo Amer­i­can.

Wes­sels says it is no sur­prise there­fore that only a mi­nor­ity of funds out-per­formed the Alsi dur­ing pe­ri­ods when the min­ing and re­sources sec­tor did par­tic­u­larly well (be­tween 2005 and 2008). But most ac­tively man­aged funds out­per­for med the Alsi when the re­sources sec­tor lagged other sec­tors (as it did from 2002 to 2004 and from 2008 to 2009).

The Swix is there­fore a fairer bench­mark for lo­cal eq­uity funds, Wes­sels says. The Swix down­weights shares that are listed on both the JSE and a for­eign stock ex­change, be­cause many of th­ese shares are held by for­eign in­vestors and are not avail­able to lo­cal in­vestors.

Nev­er­the­less, only a small por­tion of the funds re­viewed out­per­formed the Swix in most years (that is, three out of five years), es­pe­cially over the past five years.


Wes­sels also tested in­vest­ment strate­gies that se­lect funds based on their re­turns in the pre­vi­ous year. He tested the in­vest-and­hold strat­egy, as well as strate­gies that in­volved switch­ing funds through­out the seven-year pe­riod he re­viewed.

For ex­am­ple, Wes­sels tested the re­sults of in­vest­ing in funds that were in the pre­vi­ous year’s top quar­tile and switch­ing at the end of each year to funds that were then in the top quar­tile against in­vest­ing in funds with a per­for­mance that put them in the sec­ond quar­tile each year.

On av­er­age, the strat­egy of in­vest­ing in sec­ond-quar­tile funds pro­duced bet­ter re­turns than the strat­egy of chas­ing the top-per­form­ing funds.

The re­sults pro­duced by switch­ing to funds in the top quar­tile were also markedly worse than the ac­tual an­nu­alised re­turns achieved by funds in the top quar­tile be­tween 2002 and 2009. And the strat­egy only re­sulted in slightly bet­ter re­turns than those achieved by the Alsi, but the re­turns lagged the re­sults of the Swix.

No switch­ing strat­egy that Wes­sels tested pro­duced re­sults that out­per­formed the Swix.

To test the buyand-hold strat­egy, Wes­sels looked at the re­sults of in­vest­ing in and hold­ing var­i­ous types of funds based on their re­turns at the start of the pe­riod (March 31, 2002).

Funds that were in the top quar­tile at the start of the pe­riod did, on av­er­age, pro­duce the best re­turns for the en­tire seven-year pe­riod to March 31, 2009.

The re­sults were in line with the Swix but were less than what you would have earned had you been able to pre­dict and in­vest in the unit trust funds that ended the pe­riod in the top 25 per­cent of funds on per­for­mance.

Wes­sels also tested 5 000 ran­dom port­fo­lio selections for both a buy-and-hold strat­egy and one that in­volved an­nual switch­ing.

The re­sults of th­ese sim­u­la­tions were very sim­i­lar to the re­turns you would have achieved by in­vest­ing in the Alsi and, on av­er­age, less than what you would have earned from the Swix.


Wes­sels con­cludes that what­ever method­ol­ogy you use to se­lect a top-per­form­ing fund, most likely you will end up earn­ing an av­er­age re­turn over time.

Al­though some funds may have a great track record of pro­duc­ing above-av­er­age re­tur ns con­sis­tently, it is dif­fi­cult to pre­dict whether their fu­ture per­for­mance will also be con­sis­tent.

He says it is al­most as dif­fi­cult to se­lect a fund that is ca­pa­ble of out-per­form­ing the mar­ket as it is to se­lect in­di­vid­ual shares.

Pick­ing the right fund pos­si­bly in­volves more luck (as op­posed to skill) than most of us are will­ing to con­cede, he says.

You should, he sug­gests, try to con­trol what you can – namely, the cost of in­vest­ing – by in­vest­ing in a low-cost pas­sive in­vest­ment, such as an in­dex-tracking unit trust fund or an ex­change traded fund.

Wes­sels sug­gests you would do well to al­lo­cate a sig­nif­i­cant por­tion of the eq­uity com­po­nent of your port­fo­lio (25 to 50 per­cent) to a low-cost pas­sive in­vest­ment.

When it comes to the rest of the eq­uity com­po­nent of your port­fo­lio, Wes­sels sug­gests you look for an ac­tive fund man­ager whom you re­gard as a pru­dent guardian of your in­vest­ment rather than sim­ply in­vest­ing in line with the man­ager’s past track record.

He says a fund with a con­sis­tent track record may in­stil con­fi­dence in the fund man­ager you choose, but you must con­sider the man­ager’s in­vest­ment process and phi­los­o­phy.

The quar­terly unit trust per­for­mance fig­ures for pe­ri­ods up to 30 years, the PlexCrown Fund Rat­ings and the con­sis­tency of per­for­mance re­sults are pub­lished on the Per­sonal Fi­nance web­site. Go to www.pers­ and then click on the “Unit trust re­sults” link on the left-hand menu.

Newspapers in English

Newspapers from South Africa

© PressReader. All rights reserved.