...

Posts Tagged ‘Investing’

, , , , , , , , , , , ,

How Can You Tell Who Will Be a Good Investor?

In Uncategorized on December 22, 2011 at 3:06 pm

(image credit:  money.cnn.com)

We talked about what a great investor looks like, but how do we figure out who will be a good investor?  You gotta measure it to manage it, so do you look at how much money someone has made?  Their returns?  Or, more sophisticated, at an investor’s alpha or Sharpe Ratio?  These types of measures are the focus of fifty years of finance literature.  I can’t argue with their precision.  But if you care about future returns, we do know that they measure the wrong thing.

The traditional measures of investing prowess only measure the historical output.  In order to decide where to invest and create durable/improving investment performance you have to measure how the inputs make the outputs.  Just imagine sitting down with an investor for a year-end review and saying:  “Look you are doing OK, but we really need you to increase your returns in excess of the risk-free rate divided by the historical standard deviation of returns by at least 0.2 in the next year (Sharpe Ratio).”  Where would one even start with that?  It’s not actionable for an investor.  Like alpha, the Sharpe Ratio is a vanity metric – a figure that looks good for your end customers (limited partners), but has little to do with how to effectively deliver that product consistently to customers in the future.

If that wasn’t clear, consider another example, suppose GM exclusively measured their entire organization based on the finished vehicle.  In this case, vehicles would never get built because no one would know what to do.  To make the final product, you need to measure the performance of the various pieces of the production process in the ways that are under the control and are relevant to each part of the process.  People making seat covers need to be primarily measured on how many fault-free seat covers they can make per hour, rather than the JD Power rating of the entire vehicle.  If the interior, chassis, electronics, engine and assembly people are all executing well on their responsibilities, chances are they will make a great vehicle.  Not rocket science.

OK, so we don’t want to focus on the historical outputs, we want to measure how well the inputs are used, you get that.  What inputs should we measure then?  I will break them down one at a time, but to keep this short, I will give you the answer upfront and then show you how I got there.  The number one metric to measure is batting average.  Batting average, batting average, batting average.  In other words, the number of investments that an investor makes that are profitable divided by the total number of investments.  If an investor’s batting average over time is <50%, there are few ways to sustainably make money.  That is the breakpoint between success and a painful spiral.  A batting average <50% means that an investor will have to roll the dice for a few big winners and/or hope to allocate more capital to the winners than the losers.  It’s not sustainable or repeatable.

What is true about individual investors is true about investment organizations as well.  Obviously, if the aggregate investment organization’s batting average is <50%, they will need big winners or lucky capital allocation to perform.  But the batting average within the organization tells you even more.  If only a few partners have a batting average >50% or batting averages are declining among partners over time, that means that the partnership is sick.  If young analysts’ batting averages aren’t improving over time, that means the organization is not investing in training and/or recruiting.  In both cases, the great investment returns of the past are very unlikely to be repeated in the future even if the whole dollars, returns, alpha and Sharpe Ratio appear to be fine today.

It is clear that having a batting average >50% is at the core of making money investing, but is it an actionable metric?  Consider the year-end review to discuss a batting average:  “Look you are doing OK, but we really need you to increase your batting average from 40% to 60% next year.”  What do you do with that?  Well, go pick more stocks that go up than down.  What is the easiest way to do that?  Weed out the losers.  Do more of the winners.

In order to do that, you need to actually have data on when/where you win and lose.  I track my batting average by geography, market cap size, sector, industry and situation (IPO, spin-off, merger, etc…).  After a few years, an 80/20 emerges.  80% of my losses come from one particular type of investment .  So the actionable outcome is that I need to not trade those stocks or have the trading desk refuse to execute orders in those stocks.  In short, batting average is very actionable.

The readers of this post are smart, so I am sure a number of objections have started popping up in your heads.  Shouldn’t you measure batting average relative to the market?  Over what time frame should you measure the batting average?  Agreed.  So in an absolute return setting, a >50% batting average is what matters.  If you are long-only or benchmarked, then you gotta beat the batting average of the index.  If you are long-short, with certain exposure weightings, then you can benchmark against the index with those exposure weightings.  The same flexibility goes for timeframe.  If you care about performance over 3 years, then track batting average over 3 years, not year-to-year.  If you care about performance daily, then track the batting average every day.  A batting average is simple division, make it work for you.

Obviously, batting average does not = profits.  You gotta multiply the batting average times the exposure (the amount of money allocated to each investment) to get profits and we’ll go over measuring exposure in the next post.

, , , ,

View from the Bottom #24

In Uncategorized on December 13, 2011 at 1:40 am

I’m going to transition the View from the Bottom from telling you what will happen in the coming 6 months to writing about how to make more money investing.  Here’s why I am changing the focus:

The public (and private) investment business has gone from an “explorer” business where returns were so fat that anyone with a calculator, a phone and a command of the English language could make enormous sums of money – to an “exploitation” business where profits are thin and are eeked out through efficiency.  Don’t worry, this happens to every industry over time, so we know there are two implications that we can prepare for:

1)  There will be less alpha to go around.

2) Investment firms will compete on how well they are managed.  More specifically, how quickly they can discern and give capital to investors who can produce abnormally large returns.  Investment firms that can do this will attract more LP money, better people and more carry dollars to buy that house and that new pair of shoes, every now and then.  If you just thought – don’t we do that already?  You don’t.  Keep reading.

So let’s look at what investment businesses are doing now to compete on challenge #2.  Training at investment firms takes two forms, both woefully ineffective in the current environment:

1) The Osmosis Model:  A new analyst is paired with a portfolio manager and engages in an “apprenticeship”.  The analyst assists the portfolio manager in their work and learns by doing.  The observed hit rate in developing great investors in this model is low (<25%) because it assumes that analysts can tease out lessons from an extraordinarily complex set of variables simply by observation.  In addition, the portfolio manager’s and the analyst’s risk tolerance, personalities and experiences may differ and thereby greatly impede learning.

2)  The Filter Model:  Hire a bunch of analysts, give them some money to manage, fire the bottom 20% of performers every 6 months.  A few years later, you have a group of people that is either lucky, smart or both.  Again, the observed hit rate is low <25% because you have no idea who is lucky or has skill or whether the skills are appropriate for the outlook going forward.

It is easy to throw stones…so what is the better alternative?  Well, it becomes clear if you just breakdown where investment profits come from.  *ALL* investment performance, regardless of asset class, is a product of two factors:

1)  Batting average – the percentage of investment picks that are profitable vs. not profitable.

2)  Exposure – the weighting of investment picks, i.e. how much money you choose to put behind each investment.

That’s it.  Batting Average  x  Exposure = Investment Performance.  Every.  Single.  Time.  So in developing investors, your goal is really to get the team to have a batting average well-above 50% and have exposure that favorably selects the batting average.  With that, you will always make money in all market environments.  So how do you do that?  In my experience and observation, this is a result not of putting resources towards “training people”, but of figuring out what investors are already good at and amplifying those skills…which leads me to what I call the Discovery Model.

The Discovery Model:  The 1-4 year investment “training period” is far too short for someone to learn how to invest.  The only thing that can be reasonably ascertained during this period of time is what an investor is good or bad at.  In order to produce a viable batting average >50%, an investor should focus on what they are good at and stop doing what they are bad at.  This concept extends to exposure management as well.  If an investor’s exposure selection underperforms the batting average, then an investor should not select exposure, regardless of tenure.

The implication of the Discovery Model is that investment firms that can discover what their analysts are good at the fastest and allocate the greatest amount of responsibility to these skills, win.  Have you heard that in an LP meeting before?  Never, right?  You’ve probably been lulled to sleep by stories about investment process, discipline, training programs and great people instead.

This post is getting long, so I will stop here before going into detail on how to implement the Discovery Model as an individual or an organization.  Instead, I will leave you with an “Appendix” – a breakdown of my own batting average to demonstrate how this information can be used to develop analysts and wield investment research for maximum profit with minimal effort.

So my batting average over the last 6 years has been 80% and approximately 80% in each year since 2005.  This is quite high for a period where the market has gyrated between +20% and -40%.  None of the investment firms that I have worked for have gathered this data, so none of them have peeled back the data and looked to see why my batting average is so high and how it might be maintained or propagated throughout the organization.  So let’s do that now.

On the long side, my hit rate averaged about 60% during a period where the market is down.  Let’s look one step closer:  For companies that I am covering for the first time, the hit rate is about 60%, for companies I have looked at a few times before the hit rate is 80% and for companies that I have looked at for years, the hit rate is 40%.  Now that is important information for an investment organization to know.  The more I know about a company, the worse my ability to buy stocks.  So the longer I become expert in a set of companies and the more responsibility and capital I am given, the more likely it is that I lose a lot of money.  This seems strange.  But it is the reason portfolio managers and experienced investment professionals blow themselves up all the time.  The underlying cause is the human tendency to believe our own b.s.  This isn’t inherently a bad thing.  It just means, the less I invest in companies that I have gotten very comfortable with, the better I will do in any market.

On the short side, my batting average is 95%+.  Whether I have known a company for a short period of time or years, the shorts work in any market conditions.  I think this is because shorting is inherently nerve-racking and so complacency doesn’t set in.  I also have a tendency to see the worst in people and situations.  Since the point here is to become as great at investing as fast as possible, I should just do more shorts within appropriate exposures and my batting average, and likelihood of making money will go up.

What other interesting insights are there to glean from my batting average?  I have no statistical skill at determining outcomes of government regulation.  But I have a perfect batting average with IPOs on the long side in all market conditions.  This also correlates with other data that suggests when an investment landscape is undefined (i.e. there are no comparable companies, limited coverage, etc…), I am at my best on the long side.  So for me, “getting better at longs” doesn’t mean “spending more time on the long side” or “seeing the upside” or similar nonsense.  It just means focusing my efforts on buying stocks in uncertain or unusual situations.

The batting average is only half of the equation.  Selecting exposure is equally important.  I hope to cover how to gather and analyze data on both easily and systematically in the next post.

 

, , , , , ,

View from the Bottom #20

In Uncategorized on July 4, 2011 at 3:34 pm

For all the frightening headlines, the US economy is muddling along.  The sky is neither falling nor is the outlook improving.

We know the sky is not falling because we can see home prices behaving in the normal seasonal pattern of the last 50 years – rising into April, May and the summer on a month-over-month basis.  That doesn’t mean they are going up over time; it just means we don’t have a repeat of 2010 or 2008 where housing prices fell (month-over-month) throughout the summer and the government had to step in to save asset prices and bank balance sheets.

You read that the auto-related supply chain is shaky, which is true, but demand for non-auto related durable goods is just fine.  And let’s not forget that autos had earthquake disruptions and a fantastic run in demand last year.  Fantastic runs are often forgotten and incorporated into forecasts.  So considering that autos saw demand in 2010 ahead of other durable goods and 2011 is a normalizing year, autos are doing just fine.

Employment was another sore spot in May and June, but I don’t see it as much to get worked up about.  Employment data is very volatile and the 3-6 month forward indicators of employment that I put together all seem to be just fine.

The hysteria over three bad data points in one month – housing, autos and employment belies a different problem.  There is no growth outlook that holds water.  When I say the outlook is “fine” do I mean +0% or +5% like the good old days?  I mean closer to +0%.  So we find ourselves in limbo – the sky is in tact, but there is also scant hope for double-digit earnings growth without further operating margin expansion or commodity price inflation.  The markets seem to be trying to figure out if long-term earnings growth will be low single-digits or high single-digits and whether it is still appropriate to pay a double-digit P/E multiple for this kind of outlook (n.b. for all of the sticklers in the audience, I mean S&P 500 earnings growth excluding financial provisioning reversals).  Market multiples and media rhetoric are just expanding and contracting month-to-month depending on which camp is winning out.  This makes large cap beta investing hard (where most equity funds are allocated) because unless you trade these swings aggressively or know whether the S&P is going to grow earnings 3% or 8%, it is difficult to generate returns that stand out.  I don’t know how to do either, so let’s move on.

We’ll wrap it up with technology.  I get questions about the tech bubble all the time.  Is it a bubble?  Let me ask you this – how hard do people think it is to make a $1 million in tech these days?  Making $1 million should be hard.  Unless you have exceptional talents at getting money from customers or your parents have a billion dollars that they plan to give you, making a million should take about a lifetime or longer.  Are people pouring into Silicon Valley because they are planning to grind out a professional education and career making communications/entertainment products for customers over 3 to 4 decades.  That isn’t the impression that I get.  It’s not kind of like when people who barely graduated from college were answering phones on fixed income desks in 2005 for $1 million a year; it’s exactly the same.  What is puzzling about the tech bubble is that folks get so worked up about the fact that it is a bubble.  Whether a sector is in a bubble or in the 7th circle of hell (read:  financial services), there is the same rate of opportunities to learn and invest.  Despite what you read about valuations, there are very smart technology investors out there doing their best.  If there is an investor in technology companies that has a lot to teach us, it is Paul Graham.  I wanted to take the conclusion of this View from the Bottom to highlight a series of interviews he gives with technology companies in a video that can be found through this link.

Good investors are good at asking questions.  This is their craft – pushing, listening, pushing, listening until they arrive at the nugget of truth in an opportunity.  While many of the questions Graham asks are obvious, you can see how skilled he is at leaving everything in without leaving anything out.

I’d summarize his line of questioning with each entrepreneur as follows:

1)  What is this?

2)  What does it do and who uses it?

3)  What does it do and who uses it really?

4)  Let me repeat back to you what this is and what it does to make sure I clearly understand it, is that correct?

5)  What is the hardest part of getting this off the ground?

6)  Is there nothing else that is hard(er) about getting this off the ground?

7)  What is the easiest way to do the hardest part?

8)  How do you get users to discover and engage with the site over and over again.

In my view, more than half the magic comes from questions #1-#4, which most of us are too bashful to ask because we don’t want to seem ignorant.  Instead we jump in around #5 and try to show how great we are at problem solving.  When was the last time you walked through #1-#4 in a management meeting?

,

View from the Bottom #18.2

In Uncategorized on April 14, 2011 at 2:50 am

The investment world has been in the doldrums lately.  Large investment funds have been making money, but not much.  The press and investors blame their modest returns on “high correlations” in the equity markets.  This doesn’t really explain what is happening since high correlations aren’t the cause of modest returns.  The actual cause of both high correlations and modest returns is that we have reached an evolutionary wall in value investing.  Successful investment firms will have to re-orient their styles and skills in order to generate meaningful returns again.  Here’s what I mean by this:

For those not in the investment business, high correlations are blamed for low returns because high correlations imply that macro forces rather than individual stock-picking skill are driving results.  Let’s take a brief look into history to see if this makes sense.  Periods of high correlation uniformly occur when there is stress or volatility in the market.  The actual cause of this high correlation is that everyone sitting at a trading terminal is hitting the SELL (S) button as fast as they can at the same time.  So regardless of the underlying performance of a company, the company’s stock will go down because all stocks are receiving one uniform pricing command S-S-S-S-S-S.

So what about in a period of low volatility, a lack of stress, you can have high correlations then too right?  Well no, not really.  Not even during times of “quantitative easing” or monetary expansion?  No, correlations in stocks have historically gone down during those periods.  The mechanism for this is as follows.  During times of low volatility, there are few big moves in stock prices because there is a healthy push-and-pull among market participants to set prices.  Based on diverse future expectations of stock prices, traders punch in BUYs and SELLs, B-B-S-B-B-S for a given stock – generating low volatility and lower correlations.

You might say – well that may be true, but if macroeconomic forces overwhelmed individual company performance, then correlations should be high.  Perhaps, but to be more precise, correlations should only be higher because investors are all expecting that others will punch in the same series of BUYs and SELLs.  In other words, the actual cause of this isn’t the macroeconomy- it’s that investors all expect the same series of BUY and SELL orders to be entered.  Or to be really blunt.  Everyone is thinking the same.

The fact that everyone is thinking the same has little to do with the macroeconomy.  Twenty years ago, the US was in a serious macroeconomic malaise caused by a real estate bubble where government decisions would determine financial outcomes – and correlations were 75% lower than they are now.  Everyone is thinking the same in the equity markets because we share the same instant information, tools, training and worldview – or at least most incremental buyers do.  This alternative theory on why correlations are high would be supported by the general drift upwards in correlations over the last 20 years – a time period that has seen a proliferation of information, tools and the modern value investing approach (who isn’t a value investor nowadays?).

Another hint that this perspective might be right is the fact that “value” investors are going to jail.  The current tools of  the value investing trade were developed in the early 1990′s.  Financial analysis with spreadsheets, channel checks and meetings with management were the formula for making money.  If you could get access to those, there was alpha for the taking.  Reg FD put a wrinkle into this game plan and correlations in stocks have been rising almost inexorably since then.  And now we’ve hit a wall, where investors use the same tools in such a similar manner, that some funds are finding themselves bumping up against the legal limit in order to keep their LPs or egos happy.

This isn’t that crazy of an idea; this progression has happened to every investment style over time.  Before long/short hedge funds drove “value” investing to its edge, quantitative funds in the 1980′s generated alpha because they could use computers.  This edge was gently eroded away over time and quant funds have had to push into leverage and the more obscure corners of the capital markets to make meaningful returns.  Or value investors in the 1950′s could outperform because they asked for company annual reports and read them once they received them in the mail.  This is NOT to say that it was great in the good old days.  Just the opposite, all of these ways of investing seem “easy” or quaint in retrospect, but at the time, they were revolutionary.  Which is why folks made a lot of money exploiting these strategies.  What we can learn from this is that sitting in 2011, hoping that correlations come down so that value investors can make money again, is not a winning strategy.  It may pay the bills, but it is probably not a winning strategy.    In order to win, investors will need either new information sources or different tools in order to jump over the wall and start printing meaningful alpha again.

What new sources and tools are you developing?

Follow

Get every new post delivered to your Inbox.

Join 162 other followers