What is a realistic return on value investing?

[greypiggi]

I have actually been considering to switch to a more aggressive growth allocation. There is a theory from USA where instead of having such a large bond holding, I can set aside 5 years living expenses into a cash deposit / govt bond ladder. That takes up much less cash. Then the rest
gets invested for growth that reflects my age. (Late 30s). If I do this, allocation can be just 15% bond and the rest is equity.

Anyone does it this way? Big decision to make. Basically
I am wondering if I am being too safe with so much fixed income.

[corydorus]

"
2) what do u mean by nav. For eg start of 2012,
Say I had 10m invested but I put in cash throughout the year, so end of year
I added 2m. What I do now is to use 12m as denominator to calculate returns. I know if we account for time of injections, my returns are slightly better but no easy way leh..."

There is ... use XIRR.

[csl123]

Here are some comments.

1) Seems like you are in capital preservation due to higher bond allocation?
2) Possible to use NAV method to calculate fund performance.
3) Might need to have a benchmark(MSCI World or S&P500) to determine fund performance.
4) Do you have an active or passive strategy for each asset class
- For equities, you seem to be active, but not so clear on your strategy (value, business cycle)
- For bonds, seems like you are already active.

2) what do u mean by nav. For eg start of 2012,
Say I had 10m invested but I put in cash throughout the year, so end of year
I added 2m. What I do now is to use 12m as denominator to calculate returns. I know if we account for time of injections, my returns are slightly better but no easy way leh...

Let's work through an example.
Let's assume that you wish to start a fund. You have 10 dollars you wish to invest in an open ended fund (allows future investment into the fund). I also assume there is no management fee involved.

When you invest 10 dollars into your fund. You issue 10 units to yourself. So the Net asset value (NAV) of each unit is 1 dollar. You use the 10 dollars wisely, and the fund becomes 20 dollars at the end of the year. So the NAV for each unit is 2 dollars (20 dollars/ 10 units).

Assuming your relative decides to invest 4 dollar in your fund. You issue 2 unit to your relative. Notice that the NAV of each unit remains the same. (20 + 4) / ( 10 + 2) = 2 dollars per unit.


Another year has past. But this time, market stumbles and the total value of your fund drops to 12 dollars. NAV = 12 dollars / 12 units = 1 dollar. Your relative decides to cash out half of her investment in your fund. She sells 1 unit and gets back 1 dollar. Notice that NAV remains the same 11 dollars / 11 units = 1 dollar.

When you calculate return just compare NAV over the period that you required. In this example, return is 0%.

This is easier and much simpler to manage compare to xirr.

[greypiggi]

Csi123, thanks!

Let me get this straight. That means I need to calculate portfolio value all the time (no issue here) and each time I enter cash/asset into portfolio, need to recalculate nav.

Say I started year with 10m for 10units, nov then inject 2.4m. At that time portfolio value is 12m. So nav is 1.2m. So issue 2 more units. Now fund has 12 units worth 14.4m in total. Each unit worth 1.2m.

Dec comes and the portfolio has gone up to 18m. So now each unit is worth 1.5m. So for the year, portfolio returns is 50%? Slightly lower than my method now which would use 5.4m as the overall annual gain and then divide by denominator which I would use 10m + 2.4/12 to reflect the 1 mth. That gives 53%. Or another way I do is to use 5.4/12.4 which gives a more conservative and inaccurate answer.

One question, does xirr give the same 50% answer?

[fat al]

Csi123, thanks!

Let me get this straight. That means I need to calculate portfolio value all the time (no issue here) and each time I enter cash/asset into portfolio, need to recalculate nav.

Say I started year with 10m for 10units, nov then inject 2.4m. At that time portfolio value is 12m. So nav is 1.2m. So issue 2 more units. Now fund has 12 units worth 14.4m in total. Each unit worth 1.2m.

Dec comes and the portfolio has gone up to 18m. So now each unit is worth 1.5m. So for the year, portfolio returns is 50%? Slightly lower than my method now which would use 5.4m as the overall annual gain and then divide by denominator which I would use 10m + 2.4/12 to reflect the 1 mth. That gives 53%. Or another way I do is to use 5.4/12.4 which gives a more conservative and inaccurate answer.

One question, does xirr give the same 50% answer?

In principle I use the NAV method but instead of tracking units, I keep track of return data. Using your example, returns till Nov is 20%.
Dec returns is computed as 18m/(12m+2.4m)-1=25%
Full year returns = (1+20%)x(1+25%) -1= 50%

Using excel to compute XIRR will give 55% based on below inputs:
31 Dec 2013 -10m
30 Nov 2014 -2.4m
31 Dec 2014 +18m

Here's some things I adopt which you may feel useful:
1) separate bank/loan account for investment purpose
2) cashflow in/out of bank/loan account on monthly basis (data1) - not expense or trade related
3) total purchases (data2) and sales/dividends (data3) of securities on monthly basis
4) investment related expenses captured in same bank/loan account on monthly basis (data4)
5) Monthly portfolio valuation including (data5) and excluding bank account (data6)

With above, I can automatically compute four investment measures (tho not the most accurate) by doing a portfolio valuation once a month:
Return (NAV approach) of portfolio including cash and incidental expenses
Return (NAV approach) of portfolio excluding cash and expenses
Return (XIRR) of portfolio including cash/debt/expenses
Return (XIRR) of portfolio excluding cash/debt/expenses

[greypiggi]

In principle I use the NAV method but instead of tracking units, I keep track of return data. Using your example, returns till Nov is 20%.
Dec returns is computed as 18m/(12m+2.4m)-1=25%
Full year returns = (1+20%)x(1+25%) -1= 50%

Using excel to compute XIRR will give 55% based on below inputs:
31 Dec 2013 -10m
30 Nov 2014 -2.4m
31 Dec 2014 +18m

Here's some things I adopt which you may feel useful:
1) separate bank/loan account for investment purpose
2) cashflow in/out of bank/loan account on monthly basis (data1) - not expense or trade related
3) total purchases (data2) and sales (data3) of securities on monthly basis
4) investment related expenses captured in same bank/loan account on monthly basis (data4)
5) Monthly portfolio valuation including (data5) and excluding bank account (data6)

With above, I can automatically compute four investment measures (tho not the most accurate) by doing a portfolio valuation once a month:
Return (NAV approach) of portfolio including cash and incidental expenses
Return (NAV approach) of portfolio excluding cash and expenses
Return (XIRR) of portfolio including cash/debt/expenses
Return (XIRR) of portfolio excluding cash/debt/expenses

Wah...fat_al Thanks!

My spreadsheet is developed over past 3 years and it is hard to modify man... will need to spend time to figure this all out. What I do now is to measure annualized returns since the day I start the portfolio in 2011.

So what I do is track gains (realized, unrealized, dividend, coupons, transaction cost, loan cost all net out). Then I take this gain and divide by 3 different denominators. Then I annualize the gains to get annual gains since 2011 to date.

3 denominators are :
1) Net worth at start in 2011 + various big cash additions over the years.
2) Net worth at end 2013 + cash injection this year.
3) Average networth per year divided by number of years.

I know these are all approximations but the numbers are all quite close with the midpoint being (2).

However, I do also track annual returns based on eg.
2013 net gain/value of portfolio at end 2013. Again this underestimates return since I am using the highest possible denominator (in a good year).

So why am I trying to be more accurate? it is so that I can compare properly against private bank benchmarks. I need to compare apple to apple. I am doing active mgmt and if I cannot outperform PB discretionary by at least 1%, I think easier to just let them do it all. Past years so far ok. since I beat them by about 1+% and I know my method is more conservative some more. They should use NAV method right? Right now I compare my annual return and use a 100 base in 2011 to calculate where the NAV is now. THen I compare with their annual figures for balanced portfolio and calculate the NAV they got. So far beat them by about 3+% over 3 years.

[corydorus]

Using XIRR is really simple since is a built-in function in Excel. Appears you have all the transactions needed to compute them.


Date: A column on all the transact dates

Transaction: Next column will be respective transaction amount. The sign of the number will be Buy (-), Sell (+), Dividend (+), Cash Injection (-), Cash Removed (+)

Just this 2 columns of your transactions. You can do yearly, multi-year, by stock, cashless or not ...

[Ferns]

Yup, using XIRR is really easy after I picked it up from corydorus.

What I'm doing now is:
1 worksheet to track transaction history
1 worksheet to track portfolio value at the start of the year and end of the year. If there is cash injection in between, like what corydorus suggested use (-)
1 worksheet to track performance of each stock I own using the method described by cory above.

XIRR can be done like this:
"=XIRR(highlight-all-the-transactions,highlight-all-the-dates)"

[csl123]

Csi123, thanks!

Let me get this straight. That means I need to calculate portfolio value all the time (no issue here) and each time I enter cash/asset into portfolio, need to recalculate nav.

Say I started year with 10m for 10units, nov then inject 2.4m. At that time portfolio value is 12m. So nav is 1.2m. So issue 2 more units. Now fund has 12 units worth 14.4m in total. Each unit worth 1.2m.

Dec comes and the portfolio has gone up to 18m. So now each unit is worth 1.5m. So for the year, portfolio returns is 50%? Slightly lower than my method now which would use 5.4m as the overall annual gain and then divide by denominator which I would use 10m + 2.4/12 to reflect the 1 mth. That gives 53%. Or another way I do is to use 5.4/12.4 which gives a more conservative and inaccurate answer.

One question, does xirr give the same 50% answer?

Firstly, I don't really quite understand how you derive 5.4m. Wouldn't the overall gain be (18 - 10 - 2.4 = 5.6M?) in the above example?

Secondly, It will be difficult to use a denominator without normalising to a single issue unit. Hard to decide which denominator to use (10, 10 + 2.4, 10 + 2.4 * 10/12)

Lastly, if you are trying to carry out benchmarking with fund performance, you are most likely trying to establish
1) Is your fund outperforming the rest of the funds and therefore justifying your pay as a fund manager.
2) decide if you should go passive instead of actively managing the fund
Therefore an accurate benchmarking is important and relevant to decision making.

Sorry, I don't use XIRR, so can't comment much. But from my limited knowledge, XIRR is similar to IRR except that the cash inflow and outflow is at irregular intervals. It is basically a zero NPV case of all cashflows at the discount rate, which is called IRR or XIRR. I am going to guess that more than half of those who use the formula has not much idea of what the IRR or XIRR really means. Similar to value investing (buy things that you know), use tools that you can understand.

[greypiggi]

Csi123,

Yes my math wrong. Gain is 5.6m.

I am not a fund manager but managing my own money after I sold business. And I need to know how I fare vs a private bank discretionary balanced portfolio so that I can decide whether to DIY or outsource to them. While I enjoy some parts of portfolio mgmt I should outsource if I cannot beat them. Of course there is also the option of just copying their portfolio allocation and doing etfs. All this is further complicated since my allocation is not perfectly like theirs. As
shared, mine is heavier on bonds, less equity, hedge funds and i use leverage to make up the lower equity allocation.