Teh Hooi Leng calls it a day (Aggregate Asset Management)

[CityFarmer]

I have always interested on the don't and do for quantitative approach of value investing. Joel's magic formula is one good case study, but a better one is Walter & Edwin Schloss. Walter's fund should be more comparable with Aggregate Value Fund.

Walter's fund size isn't disclosed, but I estimated as US$50-100 million, which is comparable with the Aggregate Value fund's (AVF) AUM of S$100 million. Fee wise, both have no management fee, but profit sharing, 20% for AVF, and 25% for Walter. Both seems following the similar net-net approach, and aimed for long term (4-5 years min)

One major difference between them is the level of diversification. Walter invested on 60 stocks typically, and max at 100 stocks, while AVF invested on 200 stocks. Allocation on each stock is max 10% for Walter, but <2% for AVF

It is still too early to conclude AVF performance, but disclosed performance of Walter between 1956-2002 was 16% net of fee, or 21% w/ fee. Will AVF able to achieve the same performance? AVF official target is 12-13%, lower than Walter's record.

In my mind, AVF might trade the volatility with max profit, with a higher diversification than Walter's.

Any comment?

[greengiraffe]

black cat, white cat as long as can catch mice will be good cat.

I btc - value investment means sustainable cashflow and yield. Noone wants just garang guni until green giraffe.

Value Buddies means diy and save on management fees.

GG

[specuvestor]

I thought Walter Schloss is a value investor not quant? To be fair comparing to a superinvestor of Doddsville is not a good benchmark.

Over the years i think so far Renaissance Tech showed it can avoid statistical pitfalls over a cycle

Anyway Aggregate reads this forum so would be great if they could clarify the methodology and what I missed:

34X $4300= $146200 (I don't know why 34 and not 33)

Compounding effect for fixed income over long period is enormous... imagine if our CPF is not compounded.... (hope this chart doesn't fall into Roy Ngerng's hands)

Dividends last 12 months should be adjusted by inflation of 2% annual as well if that is the idea of using 5% interest rate for a "depreciated" fixed deposit capital

Not trying to nit pick but thought I'll state the obvious...... anyway an odd number like 33 or odd dates always make haters hate

http://www.valuebuddies.com/thread-3828-...l#pid95193

[CityFarmer]

I thought Walter Schloss is a value investor not quant? To be fair comparing to a superinvestor of Doddsville is not a good benchmark.

Over the years i think so far Renaissance Tech showed it can avoid statistical pitfalls over a cycle

Anyway Aggregate reads this forum so would be great if they could clarify the methodology and what I missed:

Walter Schloss is a value investor, using quantitative methodology, base on official numbers of stocks and public news flow. Similar methodology for AVF. The philosophy is value for both of them. Base on my limited understanding on quant, Both of them are not exactly quants.

The selection of benchmark, depends on the purpose. If the purpose is to benchmark against a maximum result, than benchmarking against a superinvestor, is suitable.

If the benchmarking for business purposes, than it is a different story.

[tikam tikam]

I thought Walter Schloss is a value investor not quant? To be fair comparing to a superinvestor of Doddsville is not a good benchmark.

Over the years i think so far Renaissance Tech showed it can avoid statistical pitfalls over a cycle

Anyway Aggregate reads this forum so would be great if they could clarify the methodology and what I missed:

34X $4300= $146200 (I don't know why 34 and not 33)

Compounding effect for fixed income over long period is enormous... imagine if our CPF is not compounded.... (hope this chart doesn't fall into Roy Ngerng's hands)

Dividends last 12 months should be adjusted by inflation of 2% annual as well if that is the idea of using 5% interest rate for a "depreciated" fixed deposit capital

Not trying to nit pick but thought I'll state the obvious...... anyway an odd number like 33 or odd dates always make haters hate

http://www.valuebuddies.com/thread-3828-...l#pid95193

No need to be quant to view what you're missing: http://www.universetoday.com/114821/new-...or-bicep2/

Here, BICEP2 is talking about inflation (of the Universe). No hard feelings if it is a bad joke (it is) or used in bad taste.

[CityFarmer]

No need to be quant to view what you're missing: http://www.universetoday.com/114821/new-...or-bicep2/

Here, BICEP2 is talking about inflation (of the Universe). No hard feelings if it is a bad joke (it is) or used in bad taste.

I am totally clueless on the post?? May be I am not humorous enough to catch the joke, if it is.

[nsengkia]

I have always interested on the don't and do for quantitative approach of value investing. Joel's magic formula is one good case study, but a better one is Walter & Edwin Schloss. Walter's fund should be more comparable with Aggregate Value Fund.

Walter's fund size isn't disclosed, but I estimated as US$50-100 million, which is comparable with the Aggregate Value fund's (AVF) AUM of S$100 million. Fee wise, both have no management fee, but profit sharing, 20% for AVF, and 25% for Walter. Both seems following the similar net-net approach, and aimed for long term (4-5 years min)

One major difference between them is the level of diversification. Walter invested on 60 stocks typically, and max at 100 stocks, while AVF invested on 200 stocks. Allocation on each stock is max 10% for Walter, but <2% for AVF

It is still too early to conclude AVF performance, but disclosed performance of Walter between 1956-2002 was 16% net of fee, or 21% w/ fee. Will AVF able to achieve the same performance? AVF official target is 12-13%, lower than Walter's record.

In my mind, AVF might trade the volatility with max profit, with a higher diversification than Walter's.

Any comment?

I have an interest in the quantitative approach of value investing as well (some call this smart beta).

Currently I am using mostly Greenblatt’s approach but with ideas borrowed from Research Associates Fundamental Index and James O'Shaughnessy momentum.

Compared to AVF / Walter asset based approach, I feel that the earnings based approach maybe more volatile per share in that the underlying share prices respond faster AND the overall portfolio is also more volatile financial results to financial results as earnings swing more than asset values. So portfolio turnover should be higher for the earnings based approach. I have no data to back this up though.

In my mind, AVF might trade the volatility with max profit, with a higher diversification than Walter's.

Can explain what you mean?

The presence or absence of capital gains tax will determine the portfolio turnover one can efficiently tolerate and hence indirectly the number of stocks one should have and the optimal answer to the perennial debate between buy and hold versus buy and sell.

Regarding the appropriate level of diversification, I also think that market structure plays a part in determining what is optimal. The number of stocks available (e.g. 700+ in Singapore versus thousands in the US market) will determine the number of counters one can have (an AVF like approach in Singapore alone may result in poor performance) and the size of the companies will determine how much money one can deploy (invest 10% of $100 million into a single Singapore small cap and you have to make a general takeover offer ).

For the Singapore market structure, based on the performance of my smaller and hence by necessity more concentrated SRS portfolio versus my larger and hence can have more counters cash portfolio, performance suffers if one increases diversification from 30+ shares to 70+ shares (drop of about 4% in performance). I cannot say w.r.t. volatility though as I do not track it.

The fund is one of the local ventures on quantitative approach, similar as Joel's magic formula. Base on my general exposure on quantitative approach, the expected return should be around 20%, includes fee.

Can share where you happened on the 20% figure in particular if it is based on Greenblatt or AVF approach and if it is based on the Singapore market? I would really like to hope so but so far from 2005 to date I am only getting around AVF performance before fees and that also only for my more concentrated portfolio.

[erickong]

Value investors usually generate 10-15% long term. For those >15%, its exceptional skill and talent. So - whether its focused, or diversified, Buffett, Greenblatt, Graham, Neff, Klarman, Miller or Keynes style and so on - it will still work out in the end! Some may do better over certain periods, and be overtaken in other periods and so on - but all the VALUE STYLE works!

Thank you for this forum moderator.

[CityFarmer]

I have an interest in the quantitative approach of value investing as well (some call this smart beta).

Currently I am using mostly Greenblatt’s approach but with ideas borrowed from Research Associates Fundamental Index and James O'Shaughnessy momentum.

Compared to AVF / Walter asset based approach, I feel that the earnings based approach maybe more volatile per share in that the underlying share prices respond faster AND the overall portfolio is also more volatile financial results to financial results as earnings swing more than asset values. So portfolio turnover should be higher for the earnings based approach. I have no data to back this up though.

First of all, thanks for the sharing. I am not alone

I am new to Research Associates Fundamental Index and James O'Shaughnessy momentum. I will spend some time to know them more.

Your view on the volatility between earning-base and asset-base portfolio, seems logical, but I am not so sure on that. It may be a good research topic.

In my mind, AVF might trade the volatility with max profit, with a higher diversification than Walter's.

Can explain what you mean?

The presence or absence of capital gains tax will determine the portfolio turnover one can efficiently tolerate and hence indirectly the number of stocks one should have and the optimal answer to the perennial debate between buy and hold versus buy and sell.

Regarding the appropriate level of diversification, I also think that market structure plays a part in determining what is optimal. The number of stocks available (e.g. 700+ in Singapore versus thousands in the US market) will determine the number of counters one can have (an AVF like approach in Singapore alone may result in poor performance) and the size of the companies will determine how much money one can deploy (invest 10% of $100 million into a single Singapore small cap and you have to make a general takeover offer ).

For the Singapore market structure, based on the performance of my smaller and hence by necessity more concentrated SRS portfolio versus my larger and hence can have more counters cash portfolio, performance suffers if one increases diversification from 30+ shares to 70+ shares (drop of about 4% in performance). I cannot say w.r.t. volatility though as I do not track it.

Both AVF and Walter's fund share similar philosophy and methodology. The only major difference is the level of diversification. Walter's fund performance (21% w/ fee) was 8% more than AVF's target (13% w/ fee). I reckon AVF intentionally trades maximum gain with lower volatility.

AVF isn't target only Singapore, but 5 Asia countries IIRC. I do agree the market size play an important part on level of diversification, especially with a larger AUM.

Small cap are normally out of the radar of funds, thus give us, the retail investors the opportunities.

Thanks for the sharing of data. A 4% drop of performance, from 30+ stocks to 70+ stocks, seems consistent with AVF/Walter case study.

The fund is one of the local ventures on quantitative approach, similar as Joel's magic formula. Base on my general exposure on quantitative approach, the expected return should be around 20%, includes fee.

Can share where you happened on the 20% figure in particular if it is based on Greenblatt or AVF approach and if it is based on the Singapore market? I would really like to hope so but so far from 2005 to date I am only getting around AVF performance before fees and that also only for my more concentrated portfolio.

I categories funds into 3 groups. ETF passive, quantitative, and focus approach, base on the level of skill/effort required. From various sources, (interview, article, fund disclosed info etc), the general guideline on best performance for the three groups are ~10%, ~20% and >30% over long term for reasonable sized funds. The guideline was adjusted by various factors e.g. AUM size, public vs private fund, tolerance on volatility etc.

(still learning and all comments are welcome)

[zf87]

I also practice quantitative on my little portfolio. I agree that earning-base is inherently more volatile than asset-base strategy. And I would like to recommend a book called "Quantitative value", which also introduces other quantitative factors like "quality" or "bankruptcy factor". The magic formula actually combines value and quality. The book describes for example values other than ROE, like gross profits on total asset, free cash flow on total asset maybe equally good measures.