Knowing when to sell

[Klogg]

I see what you're saying but I reckon your un-quantifiables can be quantified.

--director buying selling - plenty of websites offer that info. Just a matter of knowing whether it's important, which I don't think it is, but you may know better.

--sector - how about a 100 day ROC as %? Use the relevant sector index.

--management quality: Number of times announcements disappoint market in the last 5 years? Lincoln indicators would prob do this for you.

--macro factors should be reflected in the companies financials, so you might need not consider them??


tech is suggesting that it's good to test this. I think there may be some software that backtests on fundamental data, but I'd have to Google that.

What fundamentals to test? You couldn't go past: http://www.thepatternsite.com/Fundamentals.html

My point wasn't how much emphasis to put on them, or which factors to use, but rather that they cannot be quantified.
For example, if director sells 10% of his holdings in Company A, how much should that lower my valuation? 10, 20%? 50%?

And just as a side-note - why on earth would I use the 100day ROC index for the sector to tell me whether or not to buy or sell a company? That's like saying "I shouldn't buy Oroton Group [ORL], because as a whole they don't make money"... meanwhile, they're a great performer.
In reality, this gives me the biggest opportunity for a bargain.

P.S. I don't hold ORL - did my research too late IMO

 

[Boggo]

That being said, if BRG, or any of the companies I partly own, were to take on substantial levels of debt without a VERY good reason (and none come to mind right now), I'd be selling.

OK, in the case of BRG, any other reasons why and where (price) would you sell ?

 

[Klogg]

OK, in the case of BRG, any other reasons why and where (price) would you sell ?

Well, given the following factors:
- Huge increase in profits
- Onerous lease expense about to expire (one of the larger ones)
- Increasing market share in Asia and Nth America

My valuation has come in the range of $4.00-$4.20. I know it's a big range, but there's no way I can pinpoint an exact value.

That being said, the question I'm seeking to answer is how much above this valuation do I sell it at. Right now, the 30% (above the top-end) I used previously is seeming a bit high - but this is also why I raised the topic.
To answer your question, anything above the $4.85 would mean I sell. (Using an approximate 20% over-valuation).

Of course, any information that helps me refine that estimate is very much appreciated (And thanks to all that have contributed so far!).

 

[McLovin]

I don't know much about those two companies, but taking a look at JBH, if the Debt to Equity ratio sat consistently above 30%, which in this case it sits a fair bit above, except for 2010 (although I'd usually look at financial statements, I've used ComSec this time), I would immediately be questioning whether this is suitable for me.
Not because their interest won't be covered, but because with that sort of leveraging, a small decrease in EBIT results in substantially smaller profits than it would a company of little or no debt.

Forget about the D/E. Same store sales had been flat or falling for 18 months and margins were falling. This was, to be fair, matched by the broader industry but JBH's PE was priced for growth.

Aside from that, management had telegraphed the maximum number of stores they were going to open and they were about 75-80% done. The marginal return of the 200th store was going to be significantly less than the 10th store. Out at $20 they didn't look that great. Even at $10, there's too many things going on in the idustry and I'd pass.

 

[Gringotts Bank]

My point wasn't how much emphasis to put on them, or which factors to use, but rather that they cannot be quantified.
For example, if director sells 10% of his holdings in Company A, how much should that lower my valuation? 10, 20%? 50%?

And just as a side-note - why on earth would I use the 100day ROC index for the sector to tell me whether or not to buy or sell a company? That's like saying "I shouldn't buy Oroton Group [ORL], because as a whole they don't make money"... meanwhile, they're a great performer.
In reality, this gives me the biggest opportunity for a bargain.

P.S. I don't hold ORL - did my research too late IMO

To quantify something is to give it a figure, 10% is a figure. You'd have to backtest it to know whether it's even worth considering. If it is, then you weight it according to how much it affects performance.

First you have a concern about "what if the sector affects things?", and then you basically say that sector is of no interest. Make up your frigging mind.

By the way, your ORL example is the perfect reason why FA is inferior to TA. You're contradicting yourself.

 

[Klogg]

To quantify something is to give it a figure, 10% is a figure. You'd have to backtest it to know whether it's even worth considering. If it is, then you weight it according to how much it affects performance.

First you have a concern about "what if the sector affects things?", and then you basically say that sector is of no interest. Make up your frigging mind.

By the way, your ORL example is the perfect reason why FA is inferior to TA. You're contradicting yourself.

I know what you mean by that, but my point with macro-economic factors was more to know what is happening in the industry, rather than discounting all companies from that industry based on an index.
For example, I know that if a company gave a profit guidance at the start of year of +5%, but then the AUD appreciated 15%, logic tells me their profits will be less. How much less isn't that straight-forward, but I know it'll be less.
An index won't tell me this either.

And I'd love to be able to quantify director selling as -10%. Even if I back-tested this, I'd find that each circumstance warrants different a different value.

If things were as simple as formulating and back-testing, the Efficient Market Hypothesis would hold true.

 

[Klogg]

Forget about the D/E. Same store sales had been flat or falling for 18 months and margins were falling. This was, to be fair, matched by the broader industry but JBH's PE was priced for growth.

Aside from that, management had telegraphed the maximum number of stores they were going to open and they were about 75-80% done. The marginal return of the 200th store was going to be significantly less than the 10th store. Out at $20 they didn't look that great. Even at $10, there's too many things going on in the idustry and I'd pass.

D/E still plays a role here. But point taken, this is by far a greater indicator.

 

[McLovin]

D/E still plays a role here. But point taken, this is by far a greater indicator.

JBH doesn't have overly aggressive gearing. Their interest cover is about 9x, iirc.

It would take a lot for them to get into trouble with that amount debt.

 

[Klogg]

JBH doesn't have overly aggressive gearing. Their interest cover is about 9x, iirc.

It would take a lot for them to get into trouble with that amount debt.

Yeah, it's not the interest cover that's an issue. It's the effect on profits that any decrease in revenue causes.

For example (I'm sure you already understand this, I just doubt my ability in articulating my point), if company A and B have $100k EBIT, A has 1mil shares, B has 300k shares and 700k debt, then at 7% interest, their profit per share looks like:

Per share profit = (EBIT - Interest)/No. Shares

Company A
= (100000 - 0)/1000000
= $0.10 per share

Company B
= (100000 - 49000)/300000
= $0.17 per share

If those sales halved:
Company A
= (50000 - 0)/1000000
= $0.05 per share

Company B
= (50000 - 49000)/300000
= $0.003 per share

Now the set of numbers here don't really relate directly to JBH, but they illustrate the point I was failing to make, lol.

McLovin - may I ask what factors you look at when thinking about selling?

 

[Boggo]

I know what you mean by that, but my point with macro-economic factors was more to know what is happening in the industry, rather than discounting all companies from that industry based on an index.
For example, I know that if a company gave a profit guidance at the start of year of +5%, but then the AUD appreciated 15%, logic tells me their profits will be less. How much less isn't that straight-forward, but I know it'll be less.
An index won't tell me this either.

And I'd love to be able to quantify director selling as -10%. Even if I back-tested this, I'd find that each circumstance warrants different a different value.

If things were as simple as formulating and back-testing, the Efficient Market Hypothesis would hold true.

JBH doesn't have overly aggressive gearing. Their interest cover is about 9x, iirc.

It would take a lot for them to get into trouble with that amount debt.

Yeah, it's not the interest cover that's an issue. It's the effect on profits that any decrease in revenue causes.

For example (I'm sure you already understand this, I just doubt my ability in articulating my point), if company A and B have $100k EBIT, A has 1mil shares, B has 300k shares and 700k debt, then at 7% interest, their profit per share looks like:

Per share profit = (EBIT - Interest)/No. Shares

Company A
= (100000 - 0)/1000000
= $0.10 per share

Company B
= (100000 - 49000)/300000
= $0.17 per share

If those sales halved:
Company A
= (50000 - 0)/1000000
= $0.05 per share

Company B
= (50000 - 49000)/300000
= $0.003 per share

Now the set of numbers here don't really relate directly to JBH, but they illustrate the point I was failing to make, lol.

McLovin - may I ask what factors you look at when thinking about selling?

So from what I have established so far, if you had bought JBH in early 2009 you would still be holding it because your complex reasons/factors say its ok ?

Yes or No ?

 

[Klogg]

So from what I have established so far, if you had bought JBH in early 2009 you would still be holding it because your complex reasons/factors say its ok ?

Yes or No ?

Like I said, the company holds too much debt for me to consider it. So I would never have bought into JBH to begin with.

And just as a side note, these aren't really complex reasons/factors. They're actually quite simple ideas and you can get a very good understanding of them in Security Analysis (by Benjamin Graham).

 

[McLovin]

Yeah, it's not the interest cover that's an issue. It's the effect on profits that any decrease in revenue causes.

But the interest cover is more effective at measuring operating leverage than a balance sheet ratio like DE. A company that can generate a higher RoE should be able to carry more debt relative to equity, because equity is relatively small.

Example:

Two companies with a million each in equity and both with 30% debt to equity. Both pay 5% interest.

Company A earns $400k EBIT. It's interest cover is 26x.

Company B earns $100k EBIT. It's interest cover is 6.6x.

Both have the exact same D/E. But equal beasts they are not!

That was the point I was making with JBH, it's high RoE means it can carry more debt relative to equity. There are of a course a myriad of other factors (industry, economic cycles, Mars invading etc) but it's hard to see a company with 9x interest cover having a liquidity event.

McLovin - may I ask what factors you look at when thinking about selling?

Keynes: When the facts change, I change my opinion. What do you do sir?

I re-evaluate a couple of times/year. IME, most times there are plenty of early warning signs to get out. Without outright fraud, decent companies don't go from fantastic to woeful overnight. At least that has been my experience.

I do use a PE model for some of the more short term stuff I do. I do also use it on longer term stuff but generally I do it in paralell with DCF.

 

[Klogg]

That was the point I was making with JBH, it's high RoE means it can carry more debt relative to equity. There are of a course a myriad of other factors (industry, economic cycles, Mars invading etc) but it's hard to see a company with 9x interest cover having a liquidity event.

Yeah, fair point. I can't see a company as fundamentally strong as that having any sort of liquidity event without taking over some sort of fairly large asset and needing the $$ to fund it.

Keynes: When the facts change, I change my opinion. What do you do sir?

I re-evaluate a couple of times/year. IME, most times there are plenty of early warning signs to get out. Without outright fraud, decent companies don't go from fantastic to woeful overnight. At least that has been my experience.

Very interesting.

May I ask in regards to DCF, how much emphasis you place on this value? Do you use it in conjunction with something other than a PE model?

Thanks in advance.

 

[craft]

I have three company sell triggers

The first is the reverse of the buy reason. The discounted cash flows that could reasonably be expected by holding the company forever are less than the current price.

The second is that if the business performs worse than the assumptions embedded in my valuations. I will never lower a valuation and continue to hold.

The third is to free funds for other opportunities or indulgences. The closest to being sold according to rule one is selected.

And one system sell trigger based on my historic equity curve channel. A lot of FA is subjective –This one is for protection against myself.

 

[Klogg]

And one system sell trigger based on my historic equity curve channel. A lot of FA is subjective –This one is for protection against myself.

If you're willing to, could you please give a little more information on this? What is it in your historic equity curve channel that helps you make this decision?

Thanks!

 

[craft]

If you're willing to, could you please give a little more information on this? What is it in your historic equity curve channel that helps you make this decision?

Thanks!

A linear regression line of my historical equity curve with an envelope that touches the peak and trough. I watch both marked to market equity value and ‘allocated’ equity value in this fashion. If I make a new historical low deviation from the Regression Line, I will be re-assessing my judgement from the sidelines.

 

[McLovin]

May I ask in regards to DCF, how much emphasis you place on this value?

Quite a bit, although I really like it these days because it lets me do scenario analysis. I tend to do most of my thinking with a pen, paper and calculator.

Sorry the answer is a bit vague, but I don't know, a lot of it is internalised and I'm not sure how to quantify it.

 

[tinhat]

Klogg gave us the example of his holding of BRG. So it might be useful for people to state what they would currently use as a stop-loss according to their system/s.

Here goes for me.

Current PE is 13.74
Industry PE is 11.51

Company has had strong EPS growth and forecast EPS growth for next 18 months is double digit. Oddly, revenue is not growing nor forecast to grow FY13. So I don't know what is driving EPS growth. I would have to look further into the company to find out more about the fundamentals driving EPS growth.

Industry is consumer discretionary. Company sells internationally. Will do well if consumer discretionary spending grows.

Consensus target price (Thompson Reuters) is $4.02 (six analysts).

You bought for $3.10
Price is $4.20
Capital gain $1.10 or 35%

Consensus forecast is for FY12 dividend of 21c and FY13 dividend of 23c. Let's assume 21c fully franked. Yield on your buy price, grossed up for franking credit is 9.7% If you are confident that yield will hold up or grow over the next couple of years then, for my SMSF which has one member in pension mode, that would qualify the stock for me as an income stock.

You will be picking up the interim dividend (ex div 9/3/12) of 12.5c. Assuming you will qualify for the franking credit, grossed up that is a return of 5.7%

So in terms of deciding whether to hold for long term capital growth plus income versus realise capital gains now. You're going to be forecasting a gross dividend yield of almost 10% per annum over the next couple of years by keeping your money at risk in the stock. Contrast this to the capital gain you are sitting on of 35% plus the grossed up dividend you will receive (assuming you are eligible to claim the franking credit) that total's about 40% return.

Looking at the technical indicators...

Weekly chart shows stock in Elliot wave five (third wave up) since GFC. Price has become parabolic on the most recent wave up on the weekly chart. Stochastic oscillator has been in over-bought since late February so time to keep an eye on the stock and tighten stop loss. Weekly MACD is at an all time high. In summary the stock has had a very strong run up since the end of Jan 12.

Guppy count-back stop loss is $4.02

I calculated a Chandelier Exit using a 50 day average true range of $4.018

Note that the lowest price yesterday was $4.14

If you were to sell at a stop loss of $4.00 that would represent a capital gain of 29%

 

[Klogg]

Craft - that is very interesting, haven't even considered something like that.

McLovin - thanks. This sort of approach would suit my style very well, and I'll be doing my research today into it. (I'm a pen and paper kind of person)

tinhat - From a fundamental point of view, that mirrors my thinking very well. From what I can tell, the reason for no increased revenue in FY13, yet higher EPS, is that there are a series of onerous leases set to expire at the end of this financial year, which are fairly sizable. I think they were to the tune of 5mil or so last financial year.
And I'd also agree that this is somewhat an income stock, as it's yield is one of the top reasons I'm holding it... (probably should have mentioned that earlier).

I will also look at the TA you've used and try to make my own way to those same conclusions. And thanks for using a specific example, that's helped a lot.


Also, I'd say that the objective of this thread has been met, as I have more than enough now to go away and start researching a few techniques that many have mentioned.

Thanks all!

 

[Gringotts Bank]

Bell Direct have a backtesting engine for FA, powered by Recognia. I'd forgotten about that.

You can build your own strategies, but here's a pre-built one called GARP (Growth at Reasonable Price). It uses about 7 different pre-defined FA criteria.

Dividend Growth Rate 5 Year Average
Dividend Yield
P/E (TTM)
Price/Book Ratio
EPS Growth (5-Year Historical)
Revenue TTM
Current Ratio

No idea what the sell criteria are. If I was going to use it myself, that would be my first question.

So, it performed like this versus the Ords:


5 Year Return: -20.5% -27%
3 year Return: -20.5% 20.3%
1 year Return: -16.9% -10.8%
6 month return: 3.3% 6.8%