Why do investments underperform/outperform expected returns at purchase?

[Guest Schwab711]

I'm interested in what the title says. No need to post actual returns to share your experiences. Most folks have an idea of what the expected returns are [over a 1/3/5 time period] at the time of purchase. Do any folks record their expected returns at purchase and track how their actual performance compares to them? Has it helped improve returns for anyone? If so, how so?

 

It would be interesting to see whether there's a pattern with investments that underperformed what I expected them to do and if that pattern is present with outperforming investments.

[Mephistopheles]

Generally, what I've noticed is that mediocre or shitty businesses (SHLD being the latter) do worse than I expect at the time of purchase. Great businesses do better than expected. This is because I let low p/e values or great collection of assets on the balance sheet fool me into believing the return will be great going forward. I underestimate the headwind that a mediocre or shitty business can provide.

[Uccmal]

Generally, what I've noticed is that mediocre or shitty businesses (SHLD being the latter) do worse than I expect at the time of purchase. Great businesses do better than expected. This is because I let low p/e values or great collection of assets on the balance sheet fool me into believing the return will be great going forward. I underestimate the headwind that a mediocre or shitty business can provide.

 

This is certainly what I am noticing.  I have never tracked it specifically.  Price paid is still relevant.  From my signature I wouldn't  buy Enb, Ry, BCe, fn, or Rus at todays prices.  All were at least 20-50% lower within the  past few months.  It is nice to have a good stable of good companies to add, when there are economic related downtrends - In Canada, everything seemed to balance on the trends in oil price regardless of their relationship to the commodity or Alberta. 

[Guest Schwab711]

Generally, what I've noticed is that mediocre or shitty businesses (SHLD being the latter) do worse than I expect at the time of purchase. Great businesses do better than expected. This is because I let low p/e values or great collection of assets on the balance sheet fool me into believing the return will be great going forward. I underestimate the headwind that a mediocre or shitty business can provide.

 

This is what I was thinking. It seems like most valuation methods systematically undervalue stable companies with predictable cash flows and high-quality companies with pricing power. They also seem to overvalue highly leveraged, cyclical companies, since I very rarely see models build in a "business cycle". Next step is to prove it, which is easier said than done.

 

I have been tracking my IV, 5y E®, and 10y E® at purchase on an annual basis for my holdings so the sample size is really small. However, I'm noticing that a lot of companies I've been interested in track their E® with what seems to be a fairly high correlation. I can only guess, but I would think that investments that bank on reversion to mean or multiple expansion have significantly higher volatility relative to their E® (which makes sense intuitively).

[shhughes1116]

So I have a fairly quantitative, financial background, while my wife does not.  I've found that my best performing stuff (I.e. The stuff that outperformed my expectations) are the investment theses that I can write on a napkin and she understands.  On the flip-side, the ideas that require a slide deck to explain are almost always the crappy ideas that underperform my expected returns.

 

In practice, this has resulted in a fairly large portfolio of community banks, micro-cap real estate equities, and a few demand-pull MLPs. 

[Jurgis]

So I have a fairly quantitative, financial background, while my wife does not.  I've found that my best performing stuff (I.e. The stuff that outperformed my expectations) are the investment theses that I can write on a napkin and she understands.  On the flip-side, the ideas that require a slide deck to explain are almost always the crappy ideas that underperform my expected returns.

 

In practice, this has resulted in a fairly large portfolio of community banks, micro-cap real estate equities, and a few demand-pull MLPs.

 

Your wife understands banks and MLPs?  :o  ::)

 

I know I know. On surface, these might look simple. In reality, I don't think so. You might understand a bank. I'm sure majority of people don't really. Even a community bank. Same for MLPs.

[Jurgis]

I don't track "expected returns" vs. real returns. I think it might be great idea going forward, but it's tough to go back to past purchases and quantify expected returns without bias.

 

However, to contradict majority of people above, pretty much all my X baggers (up to 10 baggers) were crappy companies. E&Ps in 2009, busted (European) bank prefs in 2009; PFHO, WDC/STX. The other X bagger subset: high growth momo stocks - AMZN, NFLX, ISRG - smaller since I'm "value investor" and I don't buy these and even if I buy, I sell them too soon.

 

These observations might be tainted with memory bias: i.e. I might not remember remember good returns on Buffett'y steady grower businesses since they were slower and took years.

[CorpRaider]

So I have a fairly quantitative, financial background, while my wife does not.  I've found that my best performing stuff (I.e. The stuff that outperformed my expectations) are the investment theses that I can write on a napkin and she understands.  On the flip-side, the ideas that require a slide deck to explain are almost always the crappy ideas that underperform my expected returns.

 

In practice, this has resulted in a fairly large portfolio of community banks, micro-cap real estate equities, and a few demand-pull MLPs.

 

I definitely believe that.

[jmp8822]

I think it depends on the market environment. In 2008-2009 many mediocre to bad companies were trading at 80-percent+ discounts. Holding those for 5 years provided great returns compared to safer blue-chip names that might have been down 30-percent. This market isn't providing the severe dislocations that existed, which is making it tough for me, at least.  Massive volatility is nice when you have the ability to pivot.

[scorpioncapital]

Define expected return? Stock price or IRR?

Your expectation or that of the investment community?

The community's perspective will matter more than yours, your expectation will be crushed or over-rewarded.

Set your expectation very low, like 'lose everything' and you'll always be pleasantly surprised..

 

 

[Guest Schwab711]

I don't track "expected returns" vs. real returns. I think it might be great idea going forward, but it's tough to go back to past purchases and quantify expected returns without bias.

 

Completely agree! That's why I'm trying to plan what I need to track now since you really can't back-fill data. Also, we are talking such small samples of data generated over a long period of time that you really only get one chance at this.

 

 

Define expected return? Stock price or IRR?

Your expectation or that of the investment community?

The community's perspective will matter more than yours, your expectation will be crushed or over-rewarded.

Set your expectation very low, like 'lose everything' and you'll always be pleasantly surprised..

 

Thanks for the response. I'm not sure I understand the first question. An E® of any format will ultimately be equivalent when you define a time period. I think the allure of undervalued net assets is the IRR appears to be astronomical, but I have seen a few studies essentially agree on their conclusions that calculate average length of time for the value gap to be realized is ~5 years. At that point, a zero-growth, 50% margin of safety has an E® = (2^(0.2))-1 - 14.9% CAGR. That assumes a 100% probability that the gap will be realized. But true 50 cent $1s are rare to find.

 

I was originally thinking my personal expectation but it might be useful to record the expectations of others. I'm not sure how I would get a consensus/average estimate for most of the stocks I'm interested in or invested in.

 

I'm not sure I agree that the community estimate is more important than a personal investment. Presumably, the only reason any of us purchase an individual security to invest in is because we believe the market has mispriced it. Otherwise, if you are seeking market-rate returns then there are numerous securities designed to provide exactly that. I'm not sure I would invest in anything specific unless I believed my estimate was more accurate than the average or community estimate and that goes for any asset type.

 

With the low expectations, the whole point of the thread is to figure out if you can earn higher returns by moving away from that outlook and becoming more detailed.

[Jurgis]

I think scorpioncapital raises couple of valid issues:

 

The actual return is based on market price. What if market either continues to undervalue the business or overvalues it a lot? Is it useful to compare the actual return against your expected return if the actual return is determined by "unreliable" partner Mr. Market?

 

Assume you assign whatever conservative expected return to AMZN/NFLX: 5%? 10%? 20%? And then market goes and does 50% annual or 100% annual for 1-2 years. Is that information useful? It doesn't necessarily reflect the quality of your decision to buy this stock. Or you buy IBM/GM with expected return 5%? 10%? 20%? and market gives you minus 5% return for 2-4 years. Does that reflect the quality of your decision?

 

I am wondering what is the information that you get when return trounces expectations or vice versa. I guess when return is way below expectations, there is possibly a lesson there.

 

I think - like scorpioncapital alludes to - it's hard to estimate the return unless you are talking about long periods (5-10 years) and you expect the return pretty much to come from growth rate. I.e. like Buffett says, long term, the price is less important since return is pretty much based on business growth and roughly corresponds to it. If you buy for shorter term gains from undervalued to fairly/overvalued, your return may be 50% a year or it could be 10% or 0% or even minus 20% depending on market's perception (and on company's performance - which you might be able to estimate better than market's reaction). If you buy for long term hold, then market's reaction may be less important and company's performance may be way more important.

 

So the long term return comparison vs. expectations might give info about your understanding of the company. For short term return comparisons, the info is more whether you can figure out what mispricings/catalists/etc. market will correct in couple years.

 

It seems to be rather interesting exercise and might be worthwhile to do it and see what it tells you. :)

[Guest Schwab711]

I think scorpioncapital raises couple of valid issues:

 

The actual return is based on market price. What if market either continues to undervalue the business or overvalues it a lot? Is it useful to compare the actual return against your expected return if the actual return is determined by "unreliable" partner Mr. Market?

 

Assuming no dividends, buybacks, or one-time gain/loss:

Actual returns = [(1 + Multiple Expansion) * (1 + Operating return)^(t) ] ^ (1/t) -1

*Operating return can be further dissected

* Multiple expansion can be compared to peers/S&P/ect

 

Adjusting for any one-time events isn't too complicated. The multiple expansion is easy to measure so I should be able to test my original thesis against any applicable component of returns.

[Deepdive]

No one has mentioned this but in bull markets the actual returns tend to do better than the expected returns. During 2013, seems like every thesis panned out. Sector rotation, sometimes certain sectors are in favor than other and sometimes the other way around. You may have the right thesis, but the sector performs badly. For event-driven names, if your event doesn't turn out as you predicted you're going to have different results. The management also plays a key role - i.e. if you find an undervalued company and the management team is willing to buy back shares, the results tend to surprise you on the upside.

[Jurgis]

 

Assuming no dividends, buybacks, or one-time gain/loss:

Actual returns = [(1 + Multiple Expansion) * (1 + Operating return)^(t) ] ^ (1/t) -1

*Operating return can be further dissected

* Multiple expansion can be compared to peers/S&P/ect

 

Adjusting for any one-time events isn't too complicated. The multiple expansion is easy to measure so I should be able to test my original thesis against any applicable component of returns.

 

Sure you can have a multiple expansion in the formula. I don't think you can predict it though.

[vinod1]

I have been tracking the performance - both operating and price - of the companies that I follow since mid 2011. Basically I do a write up of the company I am researching - I lay out the operating performance I am expecting in my conservative case, base case and optimistic case over next 3 years and 5 years (5 years part was added only later) and also the PE or BV multiple that I expect it to trade.

 

Nearly always quality stocks (my "Exceptionals", "Compounders", "Blue Chip" categories) that I track tend to reach or beat my optimistic estimates of operating performance. PE or BV multiple most of the time exceeds my estimates by some margin.

 

My "Deep Value" category of stocks however  mostly tend to reach the base or conservative case and sometimes even undershoot on operating performance. Multiples are a mixed bag and also varies a lot more.

 

So I have learned to be a bit more optimistic on the quality companies and be positively brooding on the deep value type companies.

 

Vinod