My Investment Journey

[KnowThePast]

Congratulations, KTP. These babies... they do keep you busy don't they. I have 3 daughters myself so forum posting is on a much lower priority than before, unfortunately.

Thank you skc.

Four women against 1 :1zhelp::1zhelp::1zhelp::1zhelp:


Back on topic, an awesome reply to my concerns, thank you very much.

I think you should re-evaluate what you've done first. Why are you ditching your old strategy? Are you losing patience, losing confidence, admitting failure or really think you can do better with a different approach?

This is my biggest concern about changing strategy. Am I over reacting to short term results? I've struggled with this for a few months now. Of course, just because I am aware of a bias, doesn't mean I can overcome it. But, I am fairly sure that I am making a switch for a good reason.

- The new strategy performs better in all backtests for all periods, as well as many of its variants, eg. I don't think it is a data fluke.
- it makes good sense of why it may outperform (and why others don't always take advantage of it).

Tech/a often mentions the "beginners loop" (or something to that effect), where a new student in technical analysis change his methods every so often because the previous one didn't work out. They do back test and find new kickass indicators and re-write their trading plans and improve the expected expectancy on paper... yet in the end the new approach would fail again, and the trader goes back to square one and start all over with new plan, indicator, signals etc etc.

I am not saying that is what you are doing... but I can see the parallel. So it's up to you to ask yourself the honest and hard questions. Your portfolio underperformed XJOAI by 30%... that is not good. I bet you if you are up 30% you won't be looking to change strategy.

The only thing worse is not changing, unless, of course, your first chosen strategy was an ideal one.

But yes, constantly chasing a better thing and giving up too early are common mistakes.

If you are admitting failure... then sure, sell up and move on. But you should also lookback on why the strategy failed.

No, I don't consider my current portfolio a failure. This is about moving on to something better, but I still believe in the current strategy. My dilemna is whether I make an immediate transition or a gradual one.

If you are losing patience... then I'd say stick with it. If your initial plan is indeed a 5-year (or however long) one and you are only 2-years into it... then you are doing yourself a dis-service by ending it prematurely. May be trim a thing or two that no longer fits your criteria today. But you are not allowing the strategy to work itself out in the fullness of time.

skc, it is so helpful that you came up with a list of possible reasons. It helps me greatly to tick them off, questions my judgement and biases.

I don't think patience is an issue with me here.

If you are just losing confidence... then stick with it and change nothing. We all lose confidence in our ability at one stage or another. It's no reason to change an otherwise correct course of action.

It's hard not to lose some confidence when you are losing money.

To test it, I throw all kinds of facts and numbers at it, to bypass my emotion, but the mind is a strange thing.

I would have much preferred to make this change while I am making a profit.

If you really think you can make better returns using a different approach... I'd test that assumption again. Will you have the patience and committment to stick with it over the intended timeframe? I'd also check and make sure that the return from your new approach is going to be better than the current starting point (namely, start of year 3 of a 5 year plan).

Only you can answer these questions.

Yes, only me, of course.

But it is of great use for me to hear others' opinions and critisicm to sharpen up my thinking. Thanks again for the great post.

 

[KnowThePast]

I agree.
If the portfolio doesn't have more nett winners than nett losers
Or
Much larger winners than nett losers

It's never going to be profitable.

Regardless of whatever form of analysis you take up.

If you don't change anything you'll get similar results.

Hi tech,

galumay has been very kind to respond for me on this and he is spot on. Different strategy and different timescales.

I have always stated that I expect to outperform the index after a 3-5 year period. I am terminating the strategy early, the results are not in yet. Some lessons can be learnt from them, of course.

Firstly, I look at is whether the underperformance is within acceptable limits of the strategy. It is.
Secondly, as I've written a few times before, the biggest reason for underperformance was a decision to enter the market at a very slow rate (4% per month). But, now that I am fully invested, it will not be an issue again.

Another way to look at it that may be more familiar to you - if you developed a strategy that you believed had a positive expectancy after placing hundreds of trades, how much weight would you give to the result of the first 50 trades?

 

[KnowThePast]

Gee its nice to see this post. Some important considerations well put. I wrote basically the same thing and then deleted it because I don't have faith I can communicate well enough on these types of hard questions even though they should be asked.

Only other point I had in my deleted post was that if moving on is the right answer - then I think you probably should sell all immediately. I think being a strong hand in the market is very important if you are not going to get shaken out at the worst of times - you can't be a strong hand on something you are unwinding to move on to something you believe in more. Don't need to be a strong hand if the price is moving in your favour so maybe consider some T/A to time the exits of of any stronger stocks at the moment, but consider your reaction to possible gap breakdowns of a trend if you go that way.

Another excellent point, thanks craft.

To be completely frank, I have 2 fears.
1. Admitting failure, or appearing to be.
2. Been proven wrong, again, later.

Both of these are greatly magnified by the fact that my portfolio is on a public forum. Still, probably not as much pressure as a fund manager would face if he was to contemplate such an action.

This is starting to feel like a visit to a psychologist.

 

[KnowThePast]

...and we are back talking about trading. (Hint, this thread is in the Medium/Long Term Investing Sub Forum.)

Anyway, thats it from me. Sorry ktp for once again causing your thread to be dragged off topic!

Thanks galumay for all the support.

And no worries about off topic, it seems to happen at regular intervals every five pages or so. Perhaps I should chart it..

 

[KnowThePast]

Sold my holding in BOL, a loss of $464.85 (-21.6%).

 

[tech/a]

I didn't know you had parameters or performance stats to compare with.
When you say it's within parameters--- what are they.
Can you post up the template of the strategy and the results blueprint.

I personally haven't seen a method that is profitable which after 50 trades
Is negative.

If I had a method that after 2 yrs was negative I'd be seriously looking at my data set compared with now.
If it was only over 500 trades I'd not be trading It.
It's just a far too small sample size.

I ran my system for 7 yrs live on the net
I encourage you to keep at it. I know exactly how you feel
My method was fully published and dissected by 100s of people.
Many far more qualified than I.

Personally I find the ever extending time horizon --- hog wash
If you've got an underperforming or losing strategy it will still be under performing or losing
In 3/5/10/15 yrs time

If it can't display an edge in 2 yrs
It doesn't have one.

I'm not going to give you warm fuzzy she'll be right mate give it time rhetoric
I find in your face --- stop wasting time and change your thinking--- radically---
A far more effective tool for helping traders.

Warm and fuzzy is nice
The markets not.
Enjoy the journey---join the crowd or stand alone.

Crowds supply us stand alone types lots of $$s
Not only that they stay warm and fuzzy while it happens.

 

[KnowThePast]

If it can't display an edge in 2 yrs
It doesn't have one.

This is our main point of disagreement, that we will probably never get to agree on.

So, let's leave it at that.

 

[tech/a]

KTP

It's clear you don't have a tested method
Otherwise you'd post it up.

Your trading what you and others believe is
A logical method. Its an hypothesis an idea.
A common one.

Your results are what would be expected.

Can't see the point in you posting up your method
If when challenged you don't support your view
With the test results you say you have?

 

[Triathlete]

I have always stated that I expect to outperform the index after a 3-5 year period.


Secondly, as I've written a few times before, the biggest reason for underperformance was a decision to enter the market at a very slow rate (4% per month). But, now that I am fully invested, it will not be an issue again.

I have not looked at your stocks or when they were purchased but could some of the underperformance have been due to buying at the wrong time?

In other words what direction was the stock moving in when it was bought e.g moving up, down or sideways and was the market behind the stock at the time?

These could also be issues of underperformance.

 

[systematic]

Great discussion.

On what you should do - you seem to have the answer. If you are changing strategy, then why wouldn't you want to be invested in it fully, as soon as practicably possible (tax considerations, opportunity availability of the new plan etc)?

The question of whether to change has been interesting to read. You want to, because you think you've got something better. But you don't want to abandon what you've got, because of emotion, bias etc. Which is a great danger to be aware of. We're all subject to it, even if we know about it. Myself included.

Without details, and I admit I've not read this whole thread - I can only surmise:

- Currently investing in a deep value strategy. As long as you really are (some people say they invest in cheap stocks when they really don't)...then you know it'll all be okay in the long run. I haven't seen value do fantastically over the recent period, either. That's not a reason to abandon.

- I'm not sure whether you're looking to move into (a) a different strategy altogether (you mentioned growth and other things) that ignores value altogether....or (b) whether you are looking at quality / growth factors...but you will still stick to value stocks? Momentum/growth strategies aside...I wouldn't blame you for looking at adding in quality measures. This has been a big area for me, I've been loathe to incorporate quality...but would not blame any value investor for doing so.

For example, we know that the Piotroski method (small cap value with quality measures) works in Australia. The Aussie market has enough companies that some good (or rather, financially stable) companies do become under priced.

As I look at your list, I see only 9 of 24 that qualify as a decent stocks under Piotroski. Those 9 would be a profit of 3.2% (although that's a bit meaningless, since its hard to know what they would have been when you bought them. Still, I couldn't resist a look).

Anyway, my opinion (just an opinion)...is - that as long as you are sticking to the, "factors that work" you should be okay. If it's something esoteric (or even something, "new") - well, you'd know I'd be hesitant. Even with a 10 year solid back test, I'd be hesitant.

tech/a of course, is simply wrong. Even the name of Nick Radge can be called upon to quash the silly comment about a 2 year under performance meaning there is no edge. Most of us realise that. However, pure value investing, with its tracking error, really will test anyone.

I think any investment plan goes tweaking. Even though my philosophy on the markets, and the general factors used are fixed and always will be...minor tweaks happen as the plan evolves. Colin Nicholson talks a lot about that. But a major change definitely needs good reason. Either the current plan is wrong (for you, personally) - and that's fine. Or the new plan is definitely better (with all the robustness issues resolved).

Just some random thoughts. All opinion, of course.

 

[tech/a]

Some good comment systematic.

Notably the ever increasing time horizon is a convenient argument. You can underperform indefinitely.
In the absence of meaningful strategy test results there is no benchmark other than
The index.

Happy to be completely wrong.
I am on a daily basis.
Has served me well.
Lots of little wrongs
The odd very big right.

To quote Radge
" it's fine to be wrong it's how long you stay wrong that's important."

One of my favorites
" insanity---- doing the same thing day in and day out and --- expecting a different result.

 

[So_Cynical]

Good reply skc and I support your suggestions.

IMHO the portfolio doesn't have enough large winners and too many large losers.

Review your selection criteria. The answers you need will be there in your results if you really what to find them.

Agree - too heavily skewed to mining, services and construction / too many big losers and not enough big winners. - KTP needed to take head of the negative sector sentiment and steer away.

-----------------

And what's done is done, i would sell some losers and take the tax loss and average into some of the better losers...spread your self a bit more next time.

 

[craft]

Hi KTP

Your last update had a total figure for XSOAI since 2014 of 29.26% how do you arrive at that? More like 4ish % on my figures.

Really not sure why the discussion has diverged into the merits of the original strategy - it hasn’t met the time frame so no conclusions can be drawn. .

You can look at a portfolio return as Alpha + Beta * market return.

I would expect your deep value approach to have a beta of less than 1 and a slight positive movement in the XSO; the index that best matches your universe leaves you with a pretty flat market component. But that still leaves you with some negative alpha at this particular point in time. I’m sure this possibility doesn’t surprise you. Here’s the thing – achieving consistent negative alpha is actually as difficult as achieving consistent positive alpha and if you’re not trying for negative alpha it’s even more an awesome feat – so extrapolating as per the current snapshot is unjustified. Again I guess you understand that but it seems many don’t.

Random negative alpha at some point in time, now that’s easy to achieve, actually inevitable without being an omniscient investor/trader.

If people just stuck consistently to one method its very unlikely they would produce enough alpha – either positive or negative to differ much from the market over the long run unless they bring some real judgement or skill to the table. It’s the chopping strategies because psychology steps in at a point of negative randomness which causes people to sell up, lock in the loss and abandon the strategy – and do it repeatedly until they reach their point of financial or that psychological ruin. This is why SKC’s questions about the right reason for changing are so important. With all that said you are right in saying that the only thing worse than changing is not changing – so long as it is for the right reason and I’m not questioning your response to SKC – more just putting this post out there to counter Tech’s rubbish about needing to change because of current underperformance.

Ps
Not endorsing KTP’s deep value strategy either – don’t know enough to know if his approach to it has any alpha. But the research overall does provide evidence of value, small cap and illiquid alpha albeit hard to harvest.

 

[So_Cynical]

Really not sure why the discussion has diverged into the merits of the original strategy - it hasn’t met the time frame so no conclusions can be drawn. .

KTP's deep value strategy has proven to be a failure, with time the results will keep changing but the facts as they are clearly show that his timing/entry on many stocks has been somewhat disastrous.

I buy falling stocks! - i know that my results are much better as far as % winners losers go, KTP's timing is terrible, stock selection i wont get into, its the entry timing that has clearly disappointed....deep deep value can only be found at the bottom.

 

[DeepState]

Can I suggest that you take a look at Dimensional Fund Advisers. Their style is tilting towards small and value companies (amongst other matters). They too have had a rough time of it in recent years.

The Value style has performed well for as far back as data is available to analyse it. The rationales seem plausible enough. Importantly, for whatever reason, it can go through bad patches. If it didn't, there would likely be no premium to Value.

In order to make an informed choice, perhaps it is worth going through some stuff from Dimensional and RAFI as well to see what they have to say about the recent dry spell on this concept. From that, you can make an assessment as to whether the well has been drained or whether it is just having a rough spell.

You may draw inspiration from Pzena.

It is legitimate to question your strategy in light of outcomes. You know that the stats are unlikely to yield anything material with the sample size, so it is just a call. It is also reasonable to come up with a better idea and run with that. Whatever you decide, it is well accepted that to be a Value investor requires a lot of patience. Perhaps the idea works, but you are simply unable to tolerate the pain that must be accepted.

 

[Triathlete]

The Value style has performed well for as far back as data is available to analyse it. Whatever you decide, it is well accepted that to be a Value investor requires a lot of patience. Perhaps the idea works, but you are simply unable to tolerate the pain that must be accepted.

One thing I find hard to accept with value investing is why do you want to have your money invested in stocks that may do nothing for say 12 months or even longer before the market decides it is time to push the stock higher. Would it not be better to wait until the stock actually starts to move up before putting your money in?

 

[systematic]

One thing I find hard to accept with value investing is why do you want to have your money invested in stocks that may do nothing for say 12 months or even longer before the market decides it is time to push the stock higher. Would it not be better to wait until the stock actually starts to move up before putting your money in?

It's not the individual stock itself, it's the strategy. Value can and does, disappoint. When it performs poorly and the rest of the market is doing well (which will probably includ momentum and growth investors), value then looks really bad. Kind of like in the late 90's - when Warren Buffett was a,'has-been' and value just got left behind. Of course, after the dot com crash...

Having said that - value, like any strategy - has periods of poor performance. Momentum certainly does. Yield does. Trend following periodically has a protracted poor period as well. I've not heard of a medium-long term trend follower say otherwise. Even short-term trading has the same type of periods of underperformance. For an Australian public record of that, see Brent Penfold's (of indextrader) results in 2014. I want to meet the trader that out performs every single year over a couple of decades. Some (even on this forum) seem to give out the impression that it's possible. Well, hmmm...

Back to the above. Personal opinion: I think it's possibly values sometimes underperformance when the market is doing well that makes it look really bad, and makes it hard to trade. Trend following, for example, is going to do poorly in whipsaw conditions. But at least the trader can feel like they are in the same bad boat as everyone else. Value just makes you look silly, from time to time. That's more personal view, by the way, though it is based on reasonable assumptions founded uponfacts.

 

[Triathlete]

It's not the individual stock itself, it's the strategy. Value can and does, disappoint. When it performs poorly and the rest of the market is doing well (which will probably includ momentum and growth investors), value then looks really bad. Kind of like in the late 90's - when Warren Buffett was a,'has-been' and value just got left behind. Of course, after the dot com crash...

Having said that - value, like any strategy - has periods of poor performance. Momentum certainly does. Yield does. Trend following periodically has a protracted poor period as well. I've not heard of a medium-long term trend follower say otherwise. Even short-term trading has the same type of periods of underperformance. For an Australian public record of that, see Brent Penfold's (of indextrader) results in 2014. I want to meet the trader that out performs every single year over a couple of decades. Some (even on this forum) seem to give out the impression that it's possible. Well, hmmm...

Back to the above. Personal opinion: I think it's possibly values sometimes underperformance when the market is doing well that makes it look really bad, and makes it hard to trade. Trend following, for example, is going to do poorly in whipsaw conditions. But at least the trader can feel like they are in the same bad boat as everyone else. Value just makes you look silly, from time to time. That's more personal view, by the way, though it is based on reasonable assumptions founded uponfacts.

Thanks for your explanation systematic makes sense.

I guess I look at things a bit differently.

 

[KnowThePast]

Great discussion.

So many quality posts, all I needed to do was ask.

On what you should do - you seem to have the answer. If you are changing strategy, then why wouldn't you want to be invested in it fully, as soon as practicably possible (tax considerations, opportunity availability of the new plan etc)?

I am leaning towards selling a large portion of my losses for tax purposes.

The new strategy does not currently have enough stocks available for all my capital. So, I will keep my existing stocks, but sell them as soon as there's an opportunity in the new strategy.

- Currently investing in a deep value strategy. As long as you really are (some people say they invest in cheap stocks when they really don't)...then you know it'll all be okay in the long run. I haven't seen value do fantastically over the recent period, either. That's not a reason to abandon.

- I'm not sure whether you're looking to move into (a) a different strategy altogether (you mentioned growth and other things) that ignores value altogether....or (b) whether you are looking at quality / growth factors...but you will still stick to value stocks? Momentum/growth strategies aside...I wouldn't blame you for looking at adding in quality measures. This has been a big area for me, I've been loathe to incorporate quality...but would not blame any value investor for doing so.

For example, we know that the Piotroski method (small cap value with quality measures) works in Australia. The Aussie market has enough companies that some good (or rather, financially stable) companies do become under priced.

As I look at your list, I see only 9 of 24 that qualify as a decent stocks under Piotroski. Those 9 would be a profit of 3.2% (although that's a bit meaningless, since its hard to know what they would have been when you bought them. Still, I couldn't resist a look).

Just finished reading "Quantitative Value", which talks about this precisely. I am doing something similar, but while the book advocates automating the quality/fraud filtering, I've decided to make it a manual process.

I am also working backwards - instead of looking for value, then checking quality, my new automated strategy is about finding quality, growth, small companies, then checking for value.

Looking over backtest results of this universe, it is clear that the vast majority of winning trades, including outliers are companies that have a viable, profitable business model and a valuation that is not too excessive. I couldn't find any satisfactory way to automate this check and so decided to commit the sin of adding manual work into an automated process.

This does now mean that I will be spending more time then before going through the reports and studying the businesses. In addition to the constraints of capital and opportunity, I will be constrained by the time I have available to study new investments. Thankfully, there aren't that many that qualify in ASX.

Anyway, my opinion (just an opinion)...is - that as long as you are sticking to the, "factors that work" you should be okay. If it's something esoteric (or even something, "new") - well, you'd know I'd be hesitant. Even with a 10 year solid back test, I'd be hesitant.

It's nothing fancy or obscure, but is certainly not widely published.

"Quantitative Value" had a good chapter on the topic of data mining as well - If you run enough backtests, you will find that dairy prices in Bangladesh had a direct corelation with the daily closing price of the DOW, with 90%+ degree of confidence.

One of the ways to combat it is to always start with a theory that makes economical sense and why it will continue to have an edge. Then you backtest that theory from every angle and go through all the individual trades to check for various possible errors.

I am comfortable that I've done enough of that.

Thanks again for a great contribution.

 

[KnowThePast]

Hi KTP

Your last update had a total figure for XSOAI since 2014 of 29.26% how do you arrive at that? More like 4ish % on my figures.

I started 03 June 2013, when it was 4855. Today it is 5664, up 16.7%. I've mistyped it in my update, it seems I've switched XAO and XSO columns.

It's the first few months that killed relative performance, at a time when I had less then 20% of my portfolio invested. Starting just 4 months later, in Oct, when I switched to deep value strategy, XSOAI was 5590, for a total performance of just 1.3%.

I always include the index from the date of inception, but as I've written before, it is a really poor indication of my relative performance in the first 1-2 years.

Underperformance is there, but as you kindly picked up, it is much smaller then raw figures show.

Really not sure why the discussion has diverged into the merits of the original strategy - it hasn’t met the time frame so no conclusions can be drawn. .

You can look at a portfolio return as Alpha + Beta * market return.

I would expect your deep value approach to have a beta of less than 1 and a slight positive movement in the XSO; the index that best matches your universe leaves you with a pretty flat market component. But that still leaves you with some negative alpha at this particular point in time. I’m sure this possibility doesn’t surprise you. Here’s the thing – achieving consistent negative alpha is actually as difficult as achieving consistent positive alpha and if you’re not trying for negative alpha it’s even more an awesome feat – so extrapolating as per the current snapshot is unjustified. Again I guess you understand that but it seems many don’t.

Random negative alpha at some point in time, now that’s easy to achieve, actually inevitable without being an omniscient investor/trader.

The bolded part especially, has always been a comforting thought for me to justify the use of automated, diversified strategy.

If people just stuck consistently to one method its very unlikely they would produce enough alpha – either positive or negative to differ much from the market over the long run unless they bring some real judgement or skill to the table. It’s the chopping strategies because psychology steps in at a point of negative randomness which causes people to sell up, lock in the loss and abandon the strategy – and do it repeatedly until they reach their point of financial or that psychological ruin. This is why SKC’s questions about the right reason for changing are so important. With all that said you are right in saying that the only thing worse than changing is not changing – so long as it is for the right reason and I’m not questioning your response to SKC – more just putting this post out there to counter Tech’s rubbish about needing to change because of current underperformance.

I am well aware of this limitation and this probably was the question I was trying to ask - how do I know I am making the change for the right reasons? Hearing all the opinions, including criticism has been very helpful.

Has anyone done any research or has experience with a strategy that deliberately changes from value to growth and back? Buy when when it underperforms, sell when it is doing well, ditto for growth. Not locking yourself into a strategy, but switch between several over time, as part of a master plan.

This is not at all my current intention, but another fear (again) of mine, is that I am very likely selling at exactly the wrong time for a value strategy. And, possibly switching to growth-like strategy and a high point, to make it worse.

As I said in my post above, the current plan is to keep my holdings, selling them as soon as capital is required in the new strategy. So, for a few months at least I will get the benefits and/or losses from both strategies, while it all unwinds.