Sunday, January 26, 2014

When the hedge doesn't work

I run a hedge fund which means I am meant to hedge risk. However if you hedge away all your risk you hedge away all your returns.

So you have in your head (or computer) some model of how the world works and what things should be correlated and what should not and you make your bets accordingly.

If your model is good and your position (stock) picking is good you should outperform some market (maybe a 40% bonds, 60% equity portfolio) at lower risk than the above portfolio.

But alas the hedges are necessarily imperfect hedges. [Perfectly hedged portfolios make cash returns - which at the moment is zero before fees...]

And with imperfect hedges sometimes the hedges don't work.

It doesn't matter how famous you are or how clever you are - you still have to deal with the time when your hedges don't work. It happens to all of us. [Truly - if you find a hedge fund manager who says they have never had "model failure" then find another fund manager.]

Its in this context (and noting his massive returns last year) that I remind my readers of a Bloomberg news interview with Tepper less than two months ago.




In this interview Tepper restates his case for equity markets being cheap enough and cheap compared to (say) bonds. I don't think it is a bad case - I don't think large caps are expensive. But they are more expensive than a few years ago. [See previous Bronte comments on the state of the market here and here.] [For the record I think there is a bubble in biotech and non-telecom stocks that are purchased primarily for their dividends and English language small caps that are not financial institutions.]

The real action in the above interview happens about four and half minutes in where Tepper outlines his short case for bonds. He is not short them because he thinks that there will be inflation (in fact he doesn't think there will be). Rather he is short them as a hedge against the consequence of the Federal Reserve trying to exit QE policies.

Bluntly that hedge did not work this week and particularly on Friday.

Equity markets had their worst day in yonks. The airlines (Tepper's biggest position) were not exactly good either (as they were typically off over 4 percent - double the market).

Bond markets however had a very good day. Long bond indices rose several percent in value.

In other words Tepper was long the bad stuff, short the good stuff and the short was meant to be his hedge.

His hedge did not work. Badly. I suspect he was down more than 4 percent on the day - maybe double the equity market.

As I said every single decent manager will have days when the hedge does not work. There is no implied criticism of Tepper here. [I watched that interview because I admire him...]

And I doubt his clients would be worried either. I have never watched Appaloosa pitch to clients but I suspect the risks are accurately described and I suspect David Tepper would say to his clients that if they cannot handle a low-single-digit down day or two they should not be clients. [I know Bronte would want to dissuade such people from being clients...]

But I also know what David Tepper is feeling. At Bronte we also had the worst day in our history (although it was not as bad as I think David Tepper's day was).

First you feel a little beaten up. But at some stage you need to (seriously) examine the possibility you are wrong. And at that point the emotion turns to self-loathing. It is surprising but many of the best asset managers I know really hate themselves. It's an occupational hazard - and indeed may be a pre-requisite to being really good at this game.

That said, self-loathing is not a very productive emotion. You shouldn't just stand there like a deer blinded by headlights.

The thought pattern (and this thought pattern can be ex-ante programmed into a computer if you want) has to turn to risk management - how do we behave if we continue to be wrong? At what point should we pull-the-plug and accept that we just don't get it.

The purpose of this blog post is to explore what to do with the portfolio when the hedge is not working. That is a discussion we are having at Bronte because this week it has not been working (although not in a very serious way). And I have written this from the perspective of a stock-picking long-short equity manager. [The discussion has the same intellectual components from the perspective of a computer driven trading shop - but I can't speak to the specific detail.]

Bronte's portfolio positioning

Because I don't want to give away all the tricks of the trade I am going to describe Bronte's portfolio in an idealized position. [Day-to-day situations vary.]

We have a long book which is mostly large cap like Verizon (although includes a few disclosed mid-caps like Herbalife) which should have a beta of roughly 1. On a day-to-day basis Herbalife is our highest beta stock.

We have a short book which is full of the widest and wildest range of scum, villainy, stock promotes and general nonsense we can find. The typical short position is a biotech scam stock or something like that. Some of these are very high beta indeed and the short book has a beta of about 2.

We may typically be 130 percent long, 50 percent short - but beta-adjusted we are only a net 30-40 percent long. Day to day movements usually reflect this. On a day when the market is up 1% we are typically only up 0.3-0.4 percent. When the market goes down a percent we typically go down a similar amount. To keep the maths simple I presume an ex-ante portfolio beta of 0.4 for the rest of the discussion.

As we have been right a fair bit of the time in our specific stock picks we have accumulated returns that are way better than equity market returns even though we believe we have been taking less than equity market risks.

This week however our performance has been bad. We have tracked down roughly 150 percent of equity market returns. Friday was - as I said - the worst day in our history - but the loss was only 150 percent of equity market returns...

But whilst Friday was the worst day in our history this month has not been the worst month in our history. We have only had one month where the model looked truly broken - and that was January 2012. We wrote it up at length for our clients. In that month we went down roughly 6 percent when the markets went up roughly 5 percent.

As our day-to-day beta is roughly 0.4 we should have been up roughly 2% in a 5% up market. Our 6 percent loss was about 8% of under performance - by far the worst month we have ever had.

In January 2012 we got both ends of our portfolio wrong. Some of our longs went down in a good market. Our shorts ripped up in our face. The letter we wrote at the time noted the largish losses we were taking on Google. [Google was one of our biggest positions and it fell from $660 to $580 or so that month. As discussed in the letter we purchased a tiny amount more...]

However Google wasn't where the action was. Our shorts ripped up in that month - some costing us more than 30 percent. Ex-ante we estimated their beta wrong.

Lets model how this looks.

Our ex-ante position was

130 long,
50 short,
Total position 180 - ie gross leverage of 1.8 times.

At the end of the month our longs had gone up say 2 percent (substantially less than market).
Our shorts had gone up 18 percent (substantially more than market).

Our position would thus be:

132.6 long
59 short
adding up to 191.6

However we have had losses - on these numbers adding up to 6.4 percent. Our capital has decreased from 100 to 93.6.

Our gross leverage has thus increased to 191.6/93.6 - or roughly 205 percent.

This is kind of shocking - we had 6.4 percent of losses but our gross leverage increased by 25 percentage points.

This is a real hedge isn't working situation. We had no choice but to cut positions - and we cut them fairly aggressively. We would have loved to add more Google into that slide - we really would have loved it. But we couldn't. We were forced to cut positions - and could only add more Google because we chose to cut other positions.

--

Now lets compare this to this week. Our shorts have been going down - but our longs have been going down far more. (Witness Herbalife's big decline and Herbalife is not our only bad stock this week.)

Our aggregate position size has been shrinking more or less in line with our capital. We are not becoming dramatically more leveraged. And because we are not becoming dramatically more leveraged we can sit it out.

David Tepper's leverage probably* won't have changed by more than a couple of percent either. He can sit it out too.

And so we can examine (and will examine) individual positions but the market hasn't pointed a gun at our head and forced us to take off positions. Inaction is thus an acceptable policy.

So I can just go back to self-loathing then. I guess David Tepper can do the same (but I suspect he hides it better than me...)




John

*I am not privy to Tepper's aggregate positioning. All I know is from the above interview - hence the word "probably".

31 comments:

  1. what a crap post,... don't hide like a baby behind wind bag tepper

    ReplyDelete
  2. My really bad month was January 2012 and I don't think I was hiding behind anyone.

    This month has not been nice though.

    ReplyDelete
  3. As you said, all hedges will fail to work at some point for a variety of reasons. That being said, there is a real key issue which I keep finding the asset management industry fails - (and let me speak in generalisations) which is alpha longs and alpha shorts does not produced a hedged portfolio.

    The reason can be viewed in a number of ways:

    - Alpha longs and shorts tend to be structural fundamental positions which tend to be somewhat crowded and 'obvious'. In your case, your shorts are probably less obvious, but since stocks move much faster in a downward direction, this causes a problem as to whether your shorts become massively crowded very quickly. The easy scenario is Friday, whereby if market participants reduce gross exposure, this means long positions are sold, and short positions are covered... resulting in a massive reversal of these 'alpha' longs and shorts, again as per Friday across almost all asset classes and intra asset classes

    - The other way to view this more empirically is a hedge is something that strips out some kind of risk exposure. Typically these are viewed as market beta, value and sector etc. You actually want HIGH correlation between the hedges because that means the reliability of the px moves are higher. The problem with alpha longs and shorts even for a non-quant guy to understand is that as an example - you get massive sector mismatch. So simply looking at beta will basically be wrong in a number of scenarios.

    So what does this all mean?

    Actually putting on historically high correlation hedges when markets break down is great because correlations typically break over a short time period and as volatility decreases, correlations moves back up. So yes, you can still fix your portfolio by implementing hedges in the near time.

    Secondly and most amusingly - time arbitrage. If you are correct, they will play out over the long term and you just don't want to have bad entry points. So you basically have to, in combination with above, somehow average into more exposure. This is not straight forward as you can't keep martingale-ing into your positions. This where sometimes structuring will help - eg Ackman's swap into options for his Herbalife short.

    Good luck.

    ReplyDelete
  4. Thanks anonymous. Wonder why all my critics are anonymous these days.

    If you have looked at this blog for a couple of years you will know that I am not a slave to conventional thinking.

    --

    But I have to be aware of risk management. Jan 2012 DEMANDED changes of me. This does not - its just painful.

    J

    ReplyDelete
  5. Anonymous, that's just lazy portfolio management. You have to care about risk especially a large daily move in your pnl. Sitting around and waiting till stuff plays out over the long term does not yield the best risk adjusted returns. The trading around the edges actually makes a huge difference to returns. Also the nature of running most public equities hedge funds mean you implicitly agree to manage to mark to market (rather than private assets which miraculously perform "better" in 08 just because they aren't marked properly).

    ReplyDelete
  6. It’s more than likely that your positions are sound, but it’s also likely that your views are shared by others. That means you are in crowded trades.
    The stock market is a social game. If some fund managers provide cogent public analysis to get other people to buy after they have taken their position, they shouldn’t be surprised that they can’t control the timing and rate of withdrawal of their fellow traveler’s positions.
    Hedge funds are funded by hot money. Managers act like they are managing permanent capital, but their investors are not loss tolerant. That makes for weak hands among those position holders subject to immediate liquidity demands from their investors.
    Add to the mix that market making banks are subject to regulatory second-guessing which impacts the sizing of positions, so any market can become a mile-wide and inch-deep. Enjoy the ride…

    ReplyDelete
  7. John, just curious, why, (or maybe you do already) not use optionality in the fund as a way to hedge out some of the book. Not saying I am a market wizard but the S&P or even the RUT pullback looked like a given, not to that extent, but there seemed to be some signs of some selling coming. Why not use those indices to buy puts or something of the sort on as a way to hedge out without owning actual shares of a company?

    ReplyDelete
  8. This comment has been removed by the author.

    ReplyDelete
  9. Well John, I for one enjoyed the post and applaud your ability to be so honest and open about your thought process. I have to wonder whether the above commenters have ever managed a long/short book - great description of the frustrations that leverage can cause - after all often you may well be right in the long term but you still have to react to prices. Would that we all could avoid being slaves to the market. Unfortunately the world does not work that way.

    I'm left very curious how that 3% potential 8-bagger in March 2012 turned out. Can you share it with us?

    ReplyDelete
  10. January isn't over.

    So this is where the large short base in T note futures comes from. Bring on the squeeze.

    ReplyDelete
  11. Tepper's outside capital has a 3-year lock-up, and a large chunk of his portfolio is permanent capital. a lot more staying power than a typical l/s fund, so volatility doesn't pain him as much in that regard

    curious if you can share your thoughts on portfolio decision making in relation to capital base. it strikes me that your business analysis is long-term, while many portfolio decisions are on much shorter time frame

    ReplyDelete
  12. I think (and certainly hope) that self-loathing is not a prerequisite to being good at fund management. My partner and I are ruthless and frequent in saying to each other "well we really screwed that one up" or "we were idiots" and analyzing our mistakes. But there is no self-loathing involved. Every investor makes mistakes.

    Having come to fund management from other careers, I am more aware than most that different people are better or worse at handling different types of stress. For myself, losing money in the fund is stressful but far less so than was project management on consulting engagements with high stakes, very tight deadlines, and dozens of people to coordinate. I feel lucky to have unwittingly moved from what was, for me, the higher-stress job to the lower-stress job. I am guessing that, for most people, fund management would be the more stressful job.

    ReplyDelete
  13. Anonymous here (who said you're a slave to the market). if you really understand risk management, you wouldn't be getting into positions that forces you to tap out. a 10-20% percentage moves in the market is chump change. If you complain about the short-term nature of your funding base, then change your investors. Or else you will forever be a slave to the market. Just remember you are buying stakes in companies and not just a betting slip. Most people say they do, but very few actually live by this. sounds like most of your readers are traders. If you understand my comment, you will understand how laughable your post is.




    ReplyDelete
  14. one of your better posts. thanks for the writeup.

    ReplyDelete
  15. Thanks for posting John. I really commend and appreciate your openness and willingness to share your process with us all.

    Side-line quarterbacks are more a distraction than anything else particularly for those of us actually putting ourselves out there for discussion and reflection.

    I really liked what The PM said about alpha long and short hedging. That seems to make sense to me.

    I follow Warren Mosler and learned ALOT about how bonds work from him. I really recommend you check him out at his website here: http://moslereconomics.com/

    Unfortunately I am sorry to say that I don't see many (only a handful literally) financial professionals or politicians who understand bonds in any real meaningful or fundamentally accurate way. And that fact is, imho, literally destroying our long-term economy and public policy decisions.

    Since you're an aussie (so is my lovely wife - from Normanhurst in fact)...I'd also like to turn you on to an incredible aussie economist that works with Warren Mosler very closely. His name is Bill Mitchell, he is a truly brilliant economist. I believe he works with U of Sydney or something like that. His blog is here: http://bilbo.economicoutlook.net/blog/

    thanks again John...carry on and keep sharing!

    ReplyDelete
  16. The issue here is that your book is exposed to more than just market risk. Another way of saying that is your risk model is missing factors. You can find a list of common factors in any academic research paper, but the general point is when your book is very exposed to common risk factors, you pnl is liable to suffer serious damage. A deeper question you may or may not want to ask yourself is: are the things you think of as alpha really just systematically risk you're lucky to be escaping. That, however, is a question not many money managers can comfortably ask themselves ...

    ReplyDelete
  17. Beware the copulas for they play havoc with thy hedges! The really interesting question though is whether this indicates a phase transition in market sentiment/behaviour?

    ReplyDelete
  18. January 2012 was probably a bad month for a lot of people, as a lot of small cap, beaten down stocks rallied hugely in that month. Some people thought it short covering, or position rotation at the beginning of the year. Certainly not long enough to be a beta really like 2009Q2. I have no definitive idea.

    As you pointed out, any decent hedge fund manager will suffer from times when the hedge doesn't work. You can't hedge everything, just as you can't foresee everything. Hopefully you just see enough things correctly to make money.

    If it's anything like Aug 2007 or Jan-Feb 2012, then closing out now would be a mistake. If it's like 2009, then closing out now will preserve alpha. Question is, which scenario is more like the present? 2007 & 2012 seemed to be a short term market dynamic based on participants' changing positioning. 2009 ran for much longer, as all kinds of people re-entered the market, so that the "dislocation" or"phenomenon" ran for much longer. Hence the hedges didn't work, for much longer.

    No real insights to offer but that's my 2c worth of comment.

    ReplyDelete
  19. Isnt this worry about providing smooth returns (or hedged) one of the main handicaps to hedge fund investing? Meaning to say, oversimplifying, the reasons you are long certain stocks is for their compounding ability and the reason you are short is for their dodgy nature to reveal itself fast enough to offset the high borrow fees. To overlay on top of those drivers the additional requirement that the combination of such positions also sees a smoother fund NAV path in messy markets is likely to handicap total returns as you will be trying to make the NAV price journey smooth for investor's psychological comfort but without any good long term reason?

    ReplyDelete
  20. Great post, John!

    Friday has been a thorn as well.

    ReplyDelete
  21. Roberta Flack... One of the primary qualities of a good performance is honesty. Enjoyed the post

    ReplyDelete
  22. The thing with tepper's treasury hedge was that it was based on rising rates and likely would have worked based on his comments in the video. The only problem with that is that this downmove has nothing to do with rising rates. So it's less about a failed hedge per say and more about a faulty economic outlook. A jump in equities does not equate necessarily to a boost in the economy. Indeed in ways it should mean the exact opposite if companies were really fulfilling their most basic corporate motives!

    ReplyDelete
  23. http://online.wsj.com/article/BT-CO-20140127-713877.html

    Have a great day

    ReplyDelete
  24. http://online.wsj.com/article/BT-CO-20140127-713877.html

    Follow up on the Gotham "hit piece"

    Great Blog John.

    ReplyDelete
  25. Maybe I'm just not smart enough, but I don't really understand what Tepper was trying to do with that hedge. You say he's not worried about inflation but he's short bonds as a hedge against Fed tapering hurting his long stock portfolio. So I'm trying to imagine a scenario in which stocks go down, yields rise, and inflation remains benign. It's hard to do. If stocks are falling and yields are rising, I would expect there to be inflation. Conversely if stocks are falling and inflation is benign, I'd expects yields to be falling too. So I don't get it.

    ReplyDelete
  26. I also manage a fundamentally-based market-neutral book, though it is completely quant.

    I think some of what you see has to do with these two facts:
    1) Many managers volatility-adjust position size, and the more technically sophisticated do so on a continuous basis
    2) As vol increases, these managers will reduce exposure

    Thus, any correlation between your book and these other books will be expressed as short-term negative alpha when volatility spikes.

    My own book often has its best P/L days immediately after taking large losses on these volatility spikes, even if the market is still flat or down.


    ReplyDelete
  27. Thanks, interesting post.

    Going back to first principles - how much of this is an issue with beta (as you've described it) as a concept as opposed to only looking at the way that you've applied beta in your hedging?

    I would even extend that question more broadly to the whole CAPM. How many times does this stuff have to be falsified before it's given the flick? There's been a lot of water under the bridge since LTCM!

    Ben

    ReplyDelete
  28. John,

    You made 63% in your A$ fund last year, and your beating yourself up over a 6% loss in January.
    Give yourself a break mate! I'm sure you're investors are still extremely happy with your results to date.

    You give way too much airtime to your failures, and not enough to your successors. Anyone reading your blog alone and not looking at your performance numbers could be mislead into thinking your performance had been poor.

    NO strategy works every day, week, and month, month in month out. Substantial alpha generation requires the acceptance of periodic underperformance, I would argue. Certainly continued vigilence towards risk is essential, but I think you could justifiably ease up a bit on the self-loathing though!

    Thanks for a great post though.

    Rgds,
    LT

    ReplyDelete
  29. Appreciate the almost real time introspection and sharing especially as I meet with unhappy clients tomorrow night. Always deep thinking from John, thanks again!

    ReplyDelete