Why Hold Cash?

[ERICOPOLY]

Anyhow, in my personal life I don't hold any cash at all.  I just have a large chunk hedged with puts for liquidity guarantee.  I would rather lose it to the options market than the tax man anyhow.  The puts don't cost as much as they appear in a taxable account -- the IRS shares my losses on them.

 

This is the equivalent of just buying calls (synthetic calls).  Interesting that you consider the tax angle. I suppose that doing the synthetic call is in fact more tax efficeint?  you keep taking the tax losses and can hold the shares and not pay those taxes indefinitely?  Is that the strategy?

 

Yep, you just hold the shares and kick the unrealized capital gains on the shares down the road, writing the puts off against dividend income or whenever you take a capital gain.

 

Let's say you buy something like Berkshire, and hedge it with a put. The capital gains of the shares will compound tax deferred, there will be no pesky dividends from Berkshire, and you would eventually get a step-up in cost basis on Berkshire (eliminating all tax on them) when either you or your wife expires.  Meanwhile, every year you get this valuable put that can be used to offset capital gains or dividends elsewhere in your portfolio.

 

So Berkshire merely needs to beat the after-tax cost of the puts.  I think that's a pretty good use of money that would other wise just be sitting there in cash.  But you can pick something other than Berkshire -- it's just an example.

[SharperDingaan]

We find that cash wants to be spent, holding it requires a lot of discipline, & that there is a tendency to see just the cash weighting - & not the cash flow. Try holding hedge proceeds for upwards of 2 years!

 

We also find that it pays to maintain a negative portfolio margin, so that any cash being held - reduces your margin until you actually use it. Less pressure to spend it (on something stupid) & more caution when you buy something (as you will be leveraging).  When times are good you are usually leveraged, when they aren't so good you're usually 5-10% cash. 

 

SD

 

 

[frommi]

I understand why a fund manager or a business like berkshire or ffh holds cash, but as a single person i don`t see that reason being valid. What i need is a cashflow for my running expenses, thats the reason i love dividend paying stocks. Holding cash will always come with a cost.

[rayfinkle]

Eric-can you spell out the tax situation you were describing in your BRK example?

[randomep]

 

Well Ben Graham devotes a lot of time describing a X%:(100-X)% allocation to stocks and bonds depending on market conditions.... if the bond is short duration enough then that is almost like cash. So the master says to do it, although I am always curious as to where is his mathmatical basis......

 

Same question I have for Pabrai, where is your mathematical basis? do you have proof? or simulation results?

[giofranchi]

The problem I have with holding cash for something other than the 3 yrs. packer suggest is:

Under what situation would you invest it?  We all know people who didn't invest in 2009 because things were going to get cheaper. 

 

Al,

as far as I am concerned, it goes something like this: you have to know some businesses very well, not many businesses, just a few, but make sure in those few businesses you have developed great conviction. Then, you will be able to deploy into them your “cash reserve” at the right time, provided, of course, that you have a cash reserve!

Now, when will you see me use my cash reserve?

1) If FFH trades near today’s BVPS, that makes no sense at all,

2) If LRE trades around 1.15 – 1.2 x BVPS.

When will 1) and 2) happen? I cannot know. What I know is that I won’t be able to take advantage of 1) and 2) without some cash reserve.

 

Gio

 

[sigis]

I think people are discussing two different topics here: holding cash for liquidity reasons and holding cash so that you can take advantage of opportunities when stocks get cheaper, e.g. 2009.

 

Holding cash for liquidity reasons obviously makes sense. But I don't understand the argument of holding cash because things may get cheaper tomorrow. If I see a stock with 50% upside, it's a buy. If tomorrow everything is down by 50%, I will have a stock with a 200% upside. If some other stocks offer 300% upside now, I can rebalance into those. In any case I should be able to get out of the crash well ahead. Sure, if I could time my buying perfectly, waiting till tomorrow is better. But how would I know? And if I did, why would I hold 10% cash instead of 100%?

 

 

[giofranchi]

I think people are discussing two different topics here: holding cash for liquidity reasons and holding cash so that you can take advantage of opportunities when stocks get cheaper, e.g. 2009.

 

Holding cash for liquidity reasons obviously makes sense. But I don't understand the argument of holding cash because things may get cheaper tomorrow. If I see a stock with 50% upside, it's a buy. If tomorrow everything is down by 50%, I will have a stock with a 200% upside. If some other stocks offer 300% upside now, I can rebalance into those. In any case I should be able to get out of the crash well ahead. Sure, if I could time my buying perfectly, waiting till tomorrow is better. But how would I know? And if I did, why would I hold 10% cash instead of 100%?

 

No one of the great financial minds of the past viewed cash only for liquidity reasons. Not even Mr. Buffett (otherwise, how could he have deployed so many billions in 2008 and 2009?!). If you hold cash only for liquidity reasons, by definition the amount of cash you have will never change, no matter which opportunities arise! (It will change only if your liquidity requirements change)

 

Rebalancing is easy to say, but very difficult to do… You must be sure that your other investments are performing well at the exact time the opportunity you want to grab becomes available. And, of course, you can never be sure! If, instead, they are underperforming at the same time that opportunity becomes available, your purchasing power might be greatly diminished. You can rebalance anyway, but you won’t be able to purchase as much of that opportunity as if you were holding some cash!

 

How much cash as a strategic asset to hold? I don’t know. That is the hard part, and the part those great financial minds of the past managed so shrewdly… it is up to each one to decide!

 

Gio

 

[skanjete]

1. When you assume away downside risk (very deep losses and not recovering until after several years) then I can see why cash would drag down your returns. Maybe you should try to run with Great Depression scenario like during the 1925 to 1940 period to see if cash would still outperform.

 

2. More than the mathematics of the returns, I hold cash more due to psychological reasons. Having some cash would allow you to continue to hold on to your stocks after a 50-60% fall in stocks. If you look in the mirror and do not see Buffett (or Eric) staring back to you, then I would think holding on to some cash would be prudent for an individual investor. If you are managing money for others, it might make sense to be fully invested if you find opportunities that meet your hurdle rate as the individual investors in your fund would have separate cash allocation.

 

Vinod

 

I just tried re-doing this where the year before the market as a whole goes down.  e.g.:

 

Year 1: 15%

Year 2: 15%

Year 3: 15%

Year 4: -15%

Year 5: 15%

 

where in years 1-4 in x% cash, year 5: 100% allocated.

 

This makes the idea work a bit better.  If you assume this occurs every 5 years, with the above returns, your extra cash only has to make 45% the fifth year to match the returns.  At 6 years: 67%, at 7 years: 78%.

 

That starts to be similar to Pabrai's rules, assuming he can pull off those big returns on the down years.

 

However, note that the % cash does not matter.  If the rule holds, then you should just be 100% in cash in years 1-4 and then 100% in at the high returns on year 5.  Thus, I haven't seen anything where having a low percentage of cash is better than a high percentage, it either breaks the threshold or it doesn't.

 

Have you tried to simulate more volatile circumstances?

In your example (+15,+15,+15,-15,+15), your fully invested result after 5 years is almost 50%, actually 48,7%.

Consider the following, more volatile, series : +10, +35, +0, -50, +100. The fully invested end result after 5 years is also about 50% (48,5%).

But if you would have kept 20% cash until the crash and then be fully invested in the last year, your total result would have been 58,8%, and this with less volatility.

 

The system works better with higher volatility, and the incremental system from Pabrai is even better than this simplified example, because the last 20% of his investments won't generate the 100% from the last year, but 200% for example, due to his higher return requirement for his last 20%. This way he would end up with a total return of 78,8% after 5 years in our example.

 

Some people here suggest that holding cash is akin to trying to time the market. This is not what Pabrai suggests. He suggests to up your return requirement as you invest your cash balance. You require a higher return on your last 5 or 10% of investment than on your first 80%. Due to the volatility of the markets, this only has a indirect effect of timing the market, but timing the market in not the purpose per se.

 

In fact, the same can be said about value investing. By insisting on buying value and getting a margin of safety, you are indirectly timing the market, because there is simply more value available at the bottom of the market than at the top. Market timing here again is not a purpose, but a side effect.

[bearprowler6]

I have read this discussion with much interest...it reminds me of a true story from several years ago when I was very early on in my career....

 

A successful small business owner was being encouraged by his advisors to make "better use" of the cash he had accumulated---$2 million (a lot of money at the time).

 

They could not understand or accept why he would only invest his $2 million in 30 day t-bills earning about 4% at the time (yes it was a long while ago---mid '80's). They were frustrated by his unwillingness to invest his cash more tax-efficently or into vehicles that would allow him to achieve a higher rate of return which would of course allow him to grow his money at a faster rate.

 

At the end of one meeting during which time the advisors had offered up several suggestions for the money---the business owner reminded them of something very important ---something that put all of their advice in context---he reminded them that he was the one with the $2 million!

 

From that point forward the matter was never discussed again.

[giofranchi]

I have read this discussion with much interest...it reminds me of a true story from several years ago when I was very early on in my career....

 

A successful small business owner was being encouraged by his advisors to make "better use" of the cash he had accumulated---$2 million (a lot of money at the time).

 

They could not understand or accept why he would only invest his $2 million in 30 day t-bills earning about 4% at the time (yes it was a long while ago---mid '80's). They were frustrated by his unwillingness to invest his cash more tax-efficently or into vehicles that would allow him to achieve a higher rate of return which would of course allow him to grow his money at a faster rate.

 

At the end of one meeting during which time the advisors had offered up several suggestions for the money---the business owner reminded them of something very important ---something that put all of their advice in context---he reminded them that he was the one with the $2 million!

 

From that point forward the matter was never discussed again.

 

;D ;D ;D Very funny story! Don't mess with successful small business owners! ;)

 

Gio

[racemize]

Have you tried to simulate more volatile circumstances?

In your example (+15,+15,+15,-15,+15), your fully invested result after 5 years is almost 50%, actually 48,7%.

Consider the following, more volatile, series : +10, +35, +0, -50, +100. The fully invested end result after 5 years is also about 50% (48,5%).

But if you would have kept 20% cash until the crash and then be fully invested in the last year, your total result would have been 58,8%, and this with less volatility.

 

Well, I initially just used S&P returns, which is quite volatile.  The problem is, it doesn't work for those situations.  The only way it works is if you get significantly better returns than the market on those up years but weren't so good previously.  It is simply very hard to make up for the loss of investments prior.  I need to find annual returns back to the beginning of the 20th century though, as the S&P data set isn't big enough right now. 

 

I hesitate to just make up scenarios--let's say it works in some made up scenarios, does that mean it is a good idea?  Hard to know.  For example, your 100% return example--that's never happened in the market right?  Pabrai did it once or twice with his money (I think 2002/3 and then 2009), so maybe that's why you put it in.  I hesitate to ever assume that will happen.

 

Moreover, consider the same example you just provided--if you had kept 100% cash, and then reinvested, it would have been better.  Generally, any time a model says it is good to withhold any amount of cash, you should have withheld the entire amount.  Thus, it seems that this is almost an all or nothing proposition--either don't hold cash or only deploy cash at the bottom--any choice in between is sub-optimal.  Of course, then you have to consider what situations each one wins in.  I think holding cash only works in extreme situations (e.g., such as yours where the drop is 40%+).

 

The system works better with higher volatility, and the incremental system from Pabrai is even better than this simplified example, because the last 20% of his investments won't generate the 100% from the last year, but 200% for example, due to his higher return requirement for his last 20%. This way he would end up with a total return of 78,8% after 5 years in our example.

 

Some people here suggest that holding cash is akin to trying to time the market. This is not what Pabrai suggests. He suggests to up your return requirement as you invest your cash balance. You require a higher return on your last 5 or 10% of investment than on your first 80%. Due to the volatility of the markets, this only has a indirect effect of timing the market, but timing the market in not the purpose per se.

 

In fact, the same can be said about value investing. By insisting on buying value and getting a margin of safety, you are indirectly timing the market, because there is simply more value available at the bottom of the market than at the top. Market timing here again is not a purpose, but a side effect.

 

I'm not arguing against hurdle rates--I'm trying to figure out if the graded hurdle rate makes sense.  i.e., can we show that it is rational or not to move from 2-3x requirements to 7x requirements for the last 5%?  I'd like to compare a portfolio that does the first hurdle rate all the time and a second that does the graded requirement (or just a simple requirement) all the time to see which wins in most situations.  So far, it seems like holding cash will only work in extreme down turn situations, and they would have to happen fairly often.  I'd like to come up with something more rigorous than what I've made so far, though.

[racemize]

I have read this discussion with much interest...it reminds me of a true story from several years ago when I was very early on in my career....

 

A successful small business owner was being encouraged by his advisors to make "better use" of the cash he had accumulated---$2 million (a lot of money at the time).

 

They could not understand or accept why he would only invest his $2 million in 30 day t-bills earning about 4% at the time (yes it was a long while ago---mid '80's). They were frustrated by his unwillingness to invest his cash more tax-efficently or into vehicles that would allow him to achieve a higher rate of return which would of course allow him to grow his money at a faster rate.

 

At the end of one meeting during which time the advisors had offered up several suggestions for the money---the business owner reminded them of something very important ---something that put all of their advice in context---he reminded them that he was the one with the $2 million!

 

From that point forward the matter was never discussed again.

 

Sure, I guess, but that doesn't indicate whether it was rational or not.  Him being stubborn doesn't make any sense to me to argue one side or the other--either he had good reasons or he didn't.

[giofranchi]

So far, it seems like holding cash will only work in extreme down turn situations, and they would have to happen fairly often. 

 

Joel,

I am not talking about downturns… great opportunities might arise for very different reasons… And a general market downturn is only one of them!

A BP’s offshore plant blows up, and XOM sells off abruptly… Johnson&Johnson recalls some flawed products in its portfolio of OTC drugs, and Abbott sells off abruptly… They are only two of the examples that come to my mind right now, and that I was able to take advantage of in the recent past.

How could you predict when and why those kind of opportunities might present themselves?

Imo, it simply is not possible.

 

Gio

 

[frommi]

Holding up to 50% Cash makes sense when cash yields are near S&P500 earnings yields, because than the opportunity cost is really low. In 2000 and 2007 this was the case, in 2000 it was very extreme with 2% earnings yield vs. 6.5% cash yield. (Whats the point in holding stocks then?) In 2007 it was 5.5% earnings yield vs. 4,75% cash yield. In 1987 before the crash 10y treasury bond yields were 2% higher than earnings yields = no reason to go 100% stocks. (i have to data about cash yields here, probably nearly the same)

 

But nowadays we have the opposite case, the spread between cash yields and earnings yields are extremely high. Currently there is no reason to hold cash.

[racemize]

So far, it seems like holding cash will only work in extreme down turn situations, and they would have to happen fairly often. 

 

Joel,

I am not talking about downturns… great opportunities might arise for very different reasons… And a general market downturn is only one of them!

A BP’s offshore plant blows up, and XOM sells off abruptly… Merck is sued over some flawed drug in its portfolio of patented dugs, and Abbott sells off abruptly… They are only two of the examples that come to my mind right now, and that I was able to take advantage of in the recent past.

How could you predict when and why those kind of opportunities might present themselves?

Imo, it simply is not possible.

 

Gio

 

It is not possible, but the question is, should you turn away other good investments that meet your hurdles in order to get that opportunity?  Unless they happen very often, the answer appears to be no.  Moreover, I find it hard to believe that if one of these opportunities arose, one could not find a position that had not appreciated that could be sold for the new opportunity.  The only time that is not the case is in market downturns, when most positions have not appreciated.  Then it may start to make sense, but the opportunity has to be very large to sacrifice the gains that would have been made by not holding the cash!

[SharperDingaan]

Racemize: Remodel assuming a constant 15% leverage, & a 5% cash weighting in the recession (ie: 20% cash swing). One run with perfect hindsight, one run with changes made after one year (foggy sight), & one run assuming an average constant 5% leverage. Average the results for each year, calculate the standard deviation, then look across your 5 year time series. Use only average numbers as no one can perfectly predict.

 

You should find that for small cash weightings it does not make a lot of difference.

 

And the more volatile the market, the better & more reliable the model becomes. Simply because bigger swings are easier to forsee than small ones, & hence the greater the chance you went to a +5% cash weighting in anticipation.

 

SD

[racemize]

Holding up to 50% Cash makes sense when cash yields are near S&P500 earnings yields, because than the opportunity cost is really low. In 2000 and 2007 this was the case, in 2000 it was very extreme with 2% earnings yield vs. 6.5% cash yield. (Whats the point in holding stocks then?) In 2007 it was 5.5% earnings yield vs. 4,75% cash yield. In 1987 before the crash 10y treasury bond yields were 2% higher than earnings yields = no reason to go 100% stocks. (i have to data about cash yields here, probably nearly the same)

 

But nowadays we have the opposite case, the spread between cash yields and earnings yields are extremely high. Currently there is no reason to hold cash.

 

Well, in those situations, you would not have had opportunities that met your hurdle, so you would naturally have cash.  I'm talking about the situation where you do have opportunities exceeding your hurdle, and whether or not it makes sense to hold cash in spite of them.

[giofranchi]

It is not possible, but the question is, should you turn away other good investments that meet your hurdles in order to get that opportunity? 

 

No no! If they meet my “hurdle rate”, I always invest! It is just not so easy to meet it!! ;)

 

Moreover, I find it hard to believe that if one of these opportunities arose, one could not find a position that had not appreciated that could be sold for the new opportunity.  The only time that is not the case is in market downturns, when most positions have not appreciated.  Then it may start to make sense, but the opportunity has to be very large to sacrifice the gains that would have been made by not holding the cash!

 

Well, the fact is my portfolio many times is very concentrated… It might just happen that even in a market that keeps going up 2 or 3 positions are down at the same time… Just look at FFH and LRE today, which are by far my largest investments! :(

 

Gio

[ERICOPOLY]

Eric-can you spell out the tax situation you were describing in your BRK example?

 

It's simply that you write off the puts used to hedge the position (the expectation is that they will be worthless).

 

You don't write off the gains on the underlying shares.

 

Overall, the position will be an economic gain even though you get to take a loss from time to time on the puts.

 

Under community property laws here in the US, when either you or your spouse dies all of the capital gains held in the marriage community are forgiven via a full step-up in basis upon the death of the first spouse.  You might find yourself a billionaire, with a ton of unrealized capital gains, separated from your wife and living with another woman.  It would be better not to formalize the divorce under community property law.

[SharperDingaan]

Re 50% cash hedge:

 

You may have sold 50% of a quality cyclical at what you hope is close to the cyclical top; & are waiting for the cycle to trough again - when you will repurchase. You have no idea how long that may take, but when you act you cannot afford any restrictions on your ability. Therefore you hold either very liquid paper, or less liquid paper with a very high margin limit.

 

Assuming you buy back the same number of shares, your cash return is the realized interest earned on the proceeds + the realized short gain on your repurchase - when it occurs. None of which is really being considered here .... because the underlying premise here is that cash should always be actively invested.

 

Retaining cash for future opportunities is just a passive consideration - until you need the cash!

 

SD

[racemize]

I take back my statement about the cash % always being all or nothing.  I found a situation where the ideal cash number was 35%, interestingly. 

 

Back to the models...

[racemize]

Race, do you mind sharing the latest model?

 

Sure, but they are extremely messy, so I'm not sure anyone can understand what is going on besides me:

https://docs.google.com/spreadsheet/ccc?key=0AhTPR9eP5nWedEJBNk9ucGNxOUJzTkotQXVNMkg3OHc&usp=sharing

[vinod1]

I take back my statement about the cash % always being all or nothing.  I found a situation where the ideal cash number was 35%, interestingly. 

 

Back to the models...

 

How is that even possible? I had also assumed it would be all or nothing always. Are you using fixed income for cash?

 

Thanks

 

Vinod

[stahleyp]

I take back my statement about the cash % always being all or nothing.  I found a situation where the ideal cash number was 35%, interestingly. 

 

Back to the models...

 

Speaking of which, I think that's roughly how much Klarman usually holds.