In Quest for Digital Gold

How did gold become gold? And, more importantly, why is gold staying gold?

Silver had its run during some periods and in some cultures, but silver has become “tarnished” as a store of value.

Gold meanwhile has survived the rise of fiat and major economies moving away from the gold standard. It has also survived the loss of its utility as a convenient medium of exchange.

Certainly, gold has demand from the jewelry industry and some industrial utility. But this “natural” demand is not even a close match for the existing above-ground supply. Imagine what would happen to the price of gold if all the metal in the vaults (owned by central banks or other hoarders) were to be sold.

And yet the value holds. Significant fluctuations for sure, but in the long run, it remains stable.

I am not an expert on the history of gold. To the best of my understanding, precious metals were chosen as a medium of exchange because they do not oxidize easily and remain in an uncorrupted form. They can also be traded purely by weight and are preferable to gems, which have value heavily dependent on their shape and quality.

Amidst precious metals, however, gold is neither the rarest (platinum and palladium are harder to come by) nor quite common, like silver. Gold has solid but by no means exclusive, industrial or jewelry related demand.

My guess is that over many centuries, the precious metal has hit it’s “golden” mean; it is common enough to be a worldwide currency, but not so common that mining can disrupt its price. It is useful and decorative enough to support some underlying value, but the supply is not so tight that marginal swings in demand would create wild price fluctuations.

My assumption is that through those subtle advantages, gold has simply outlasted its competition. And, as a result, it has gained so much credibility over millennia that its value can no longer be diluted or displaced.  Today, gold (XAU = $1,275) is more expensive than platinum (XPT = $915) which is much scarcer and as useful in an industrial sense.

That said, you may have guessed by the title that this post is really not about gold, but about digital assets. Until recently, I was thinking of the dichotomy between the precedence of bitcoin and the advantages of alternative protocols. I am realizing that this dichotomy may be false.

If a universal digital store of value (“digital gold”) is to be established, it doesn’t by any means have to be, and not even likely to be, the universal medium of transactions; a fortiori, it is not likely to be a universal conduit for transactions.

Bitcoin is battling to become digital gold, while ether, for example, is elbowing to become digital copper. If the analogy holds, the price of ether will be driven by technology demand, while the price of bitcoin by the need to store liquidity. Will bitcoin cash be digital silver?

As a corollary, the rise of altcoins geared towards specific purposes is not at all dilutive, but actually supportive of bitcoin. While diminishing its share of trading volume in actual transaction flows, it shields bitcoin from being demand driven and makes it a more attractive store of value.

Now, to assess the future value of bitcoin we need to answer three questions:

I.  Is there a need for digital gold?
II. Is bitcoin more like gold or more like fiat?
III. Why bitcoin specifically?

I am not going to spend time on question one; I assume that my readers appreciate the advantage of having digital value storage and I have nothing to add to the already existing discussion.

Regarding, question two, we need to address the issues of altcoins and forks. I have already explained why I think altcoins are no more dilutive to digital gold than platinum or palladium are dilutive to physical gold. And if physical mining for gold is alive and well, so will be the digital mining for digital gold.

Forks are a trickier issue because they cause many to liken a crypto-asset to fiat, and the community of miners to central bankers. So, are forks inflationary (and for the purposes of this post I use the example of recent bitcoin hard fork)?

One definition of inflation is “the tax on capital”. It describes what happens to the purchasing power of your asset if it just sits there doing nothing. Therein lies the distinction between bankers printing more money and the blockchain forking. New fiat does not go proportionately to the holders of existing money; rather it goes where the government directs it (some cynics seem to think it tends to end directed to the richest of us) via specific asset purchases such as QE or via targeted fiscal stimulus such as tax cuts or increased spending. While those things are happening, your savings may indeed be eroded.

But the hard fork gave each current holder an extra asset. Theoretically, it is a zero-sum game, but like just like stocks tend to react positively to a stock split, this particular fork went well for the asset owners. For what it’s worth, I think the fork was more like a dividend than like split, but this is not the most relevant debate here.

I think it is obvious that forks do not TAX the capital of current holders. A more interesting question, however, is posed by forward contracts. If I promised you delivery of a bitcoin on July 1st, 2017 or September 1st, 2017 this contract may have had a different value depending on how it was written.  Forward stock value, for example, dips down on the ex-dividend date. From my perspective, therefore, it is not so much an issue of forks being inflationary, but the necessity to program blockchain contracts carefully.

If the government one day announced that every dollar is to be exchanged overnight into ten dollars it would not be inflationary in itself because the purchasing power of existing money wouldn’t change. But it would be important to know what happens to existing contracts and liabilities. If you have a salary or pension guarantee, does it go up 10 times? What about debt payments? If it is all done proportionately, there is no economic impact. But if it’s done selectively a wealth redistribution could happen.

Lastly, why bitcoin specifically? It has been discussed ad nauseam why bitcoin mining qualities and limited supply gives it aspects similar to gold. But its unique advantage with respect to other crypto-assets is that was there first. And now it is in the process of outlasting the competition.

When bitcoin first started trading, I was mostly unaware and fairly agnostic of its value. As a trader, I became interested in its vertical rise in 2013 which was followed by a bear market in 2014. Notably, its drop found support; it didn’t continue to fall to permanent obscurity below the event horizon.  Instead, it stabilized, put a solid double-bottom in 2015, and started to creep up.

XBT, Oct 2012-Oct 2016


This trading pattern is consistent with precious metal behavior, only compressed to a shorter horizon.  For example, look at the slow consolidation in gold after the spike of 1980.

XAU, 1975 – Present


Or, more recently palladium after the spike of 2001.

XPD, 1995 – Present


Given the historical trading pattern in precious metals, the buy of bitcoin in 2015 was relatively easy.  But now, at 20X (including fork) the early prices, it is a more complicated trading dilemma. Bitcoin may or may not be a “bubble”, but it is important to remember that it did have a major “burst” in 2014 and, even this year, the market has witnessed bitcoin endure two meaningful corrections.

XBT, 2017 (Actual price swings were even wider than BBG-sourced data below)


And every time it survives and stabilizes, it gets extra “gold cred”.

In summary, despite the rise and fall of various alt coins and fork considerations, “time” works FOR bitcoin, not AGAINST it.  Every day it doesn’t disappear, it gets one step closer to a permanent status of digital gold.

Good luck!

Treasury Futures Trading Primer

US Government Bonds market has been one of the best games in town over the last couple of years. Especially, if you adjust for the currency. Indeed, though the yields of Japanese and European bonds have declined (priced rose) more than those of Treasuries, the currencies of the respective countries have declined against the dollar. Thus international investors would have done better staying in US Bonds.

When what is supposed to be the safest investment of the world also outperforms most world’s risky markets, it is bound to draw attention from new ranks of investors.

As one of the most vocal “bond and dollar bulls”, I’ve been getting a number of questions regarding the interest rates markets both directly from my friends and via social media.

A long-time friend with a solid academic background asked a particularly interesting set of questions about trading Treasury Futures. In this post I will give our dialog almost verbatim, as I think it might be helpful to some novice traders and even contain some elements useful for more experienced ones. I will also insert some helpful questions and answers from my blog.

To preface the dialog I want to point out a few things.

  • Treasury futures might be awkward to trade for small investors as the unit contract has a $100,000 notional.
  • The actual mechanics of those futures (with convergence factors and cheapest-to-deliver) is rather complex. This post is not meant to teach this mechanics.
  • The advantages of trading Treasury futures are the price transparency and the level field. Through futures we can all enjoy the same economics of funding and carry, as any big bank, without hidden costs.
  • Finally, keep in mind, everything below is discussion points and not trade recommendations.

Email I

Could I again ask for your help with (hopefully, rather basic) questions about fixed-income trading? I am learning to trade treasury futures, and my questions are about the underlying math.

Q: The first question is about the cheapest-to-deliver bonds for futures contracts. Am I right in my understanding that, when yields are below 6%, the cheapest-to-deliver bonds are the shortest duration bonds (among those allowed by the contract spec)? For example, the cheapest-to-deliver bonds for the classic T-Bond futures would be 15-year bonds, the cheapest-to-deliver for the 10-Year T-Note futures would be 6.5-year bonds, and so on? If I am actually wrong, how do I figure out the cheapest-to-deliver bonds for each contract?

A: First, I want to be sure you understand modified duration – it appears that you do.

Duration = – d price / d yield

Modified Duration = Duration / Dirty Price   

When the prevailing yields are (like now) way below 6%, the cheapest-to-deliver is usually the shortest maturity allowable bond. But what really matters is modified duration, so in rare cases, a slightly longer bond with much higher coupon could have a shorter modified duration and be cheaper to deliver. Current cheapest-to-deliver for every contract can be looked up.

Q: The second question is about durations. How can I figure out the duration for a specific futures contract? For example, would it be right to assume that classic T-Bond futures are roughly equivalent to a 15-year bond with 6% coupon, and thus have duration of about 11 years?

A: In this rate environment the futures contract will essentially trade with cheapest-to-deliver (adjusted by the conversion factor). There will be some small technical factors, like the carry between the spot date and the delivery date, but they are not significant. However, it is incorrect to assume bond futures will always trade like a 15-year bond. In fact, currently they trade more like 20-year bond, because there do not exist any bonds with 15-19 year maturity. Those would have been the 30-year bonds issued during the Clinton era, when the long bonds were temporarily discontinued.

Q: Suppose I expect yields to rise (over the next few weeks or months), and I would like to take a position that bets on the rise of yields. Would it be reasonable to get a spread between 5-year and 10-year T-Note futures, that is, to sell short 10-years and buy an equal amount of 5-years? Are there better ways to bet on the rise of interest rates?

A:  Betting on rising yields has been a bad idea over the last 30 years and probably still is. Google “One Chart To Rule Them All” – my twitter post appears on the first page. Check it out.

Typically selling one thing and buying another just increases the amount things that can go wrong, so I almost never recommend such strategy to express a simple directional view. If you have a view on a specific forward portion of the yield curve, this could be a correct trade to express it.

Relative direction of interest rates is rather tricky. If the Fed were to start tightening soon, 5-year futures could go down, while 10-year futures not move at all. You could lose money, despite possibly having a correct view. Of course, the opposite could happen and you could make money on both sides.

Overall I would need to know your reasoning behind the bet to have an opinion whether the strategy is appropriate.

Email II

I have two follow-up questions.

Q: You mentioned that the cheapest-to-deliver for every contract may be looked up. Could you let me know where? Probably a trivial question, but I could not find these data.

A: To look up Cheapest-to-Deliver and risk ratios:

Q: You indicated that using a spread (buying one thing and selling another) is unnecessarily complex. Is there a simpler/better way to bet on rising yields? (The first thing that comes to mind is simply shorting, say, 10-Years, but I am concerned that doing so would be “swimming against the tide”, since I would be effectively paying the coupon on the shorted bonds.)

[From my blog

Q: When you say Long Bonds will carry if the Fed stays put, could you explain in simple terms what you mean ?

A: By carry I mean the difference between the yield on a bond and the rate at which a leveraged trader can borrow money overnight, using the bond as a collateral. In the current rate environment this borrowing (funding) rate for treasury bonds is virtually zero. If the Fed were to start raising rates, the funding cost would go up, making the carry trade less profitable.]

A: Getting a “free lunch” is not easy. When you do 5/10 spread, it might appear you are avoiding the negative carry, but actually you aren’t. The 5 year forward 5 year rate slides down rapidly as it goes to 4 year forward. The nature of betting against the bond market typically leads to negative carry.

Q: Also, here is an explanation of my reasoning behind my bet on the rising yields. I realize that my reasoning is naive–I am just learning the treasuries–but it may help you understand where my questions are coming from.

I expect that yields will rise at least a little–maybe by something between 0.25 and 0.40–in the next few months, and that, if they fall, they will not fall below the Jan. 30 minimum, and not for long. (From your email, I understood that you feel far less “bullish” about the rates–which is depressing. I really hate when they are that low.)

Given that expectation, I was trying to figure out some position that would allow me to bet on rising yields without “swimming against the tide”, i.e. without simply taking a short position and paying coupon on it.

First, I drastically reduced the duration of my bond portfolio, from about 12 years to about 5. This change was done in my positions in bond mutual funds and actual treasury bonds (not in treasury futures, which I have never traded before). Now, I am trying to to a similar thing with futures, mostly for thepurpose of learning how to use them, with small positions. I first considered doing a spread between Ultra T-Bond and Classic T-Bond, but even a single-contract position turned out too large for me (in terms of potential losses and margin requirements), so instead I am looking at a spread between 10-Year and 5-Year.

A: First – terminology: being bullish on rates means expecting rates to go DOWN (price language as opposed to funding language). So I am actually bullish, but only on the long-dated rates – I have no opinion on the short end.

[ From my blog

Q: Won’t Long Bonds fall if the Fed hikes rates ?

A: Over the last decade we have found out that the Fed cannot reliably control long-term interest rates with their overnight rate policy. In fact, when the Fed tightens, the market starts projecting lower inflation and slower economic growth for years in the future. Thus, long-dated rates often don’t move or move lower early in a hiking cycle.]

If you go deep to the reason why you want to make this bet, you might find ways to do this, without swimming against the tide – this is what Part III of my book is about.

For example,

  • If you think the inflation is going to be high, you might want to buy gold.
  • If you think the economy is going to do well – you can buy an established stock such as IBM with good dividend, so you will betting on strong economy and earning carry.

Finally, if you already have a preference for higher rates – remember by actually betting on higher rates, you are doubling down.

Image: ‘Fit to Fight: Combatives Course teaches Airmen the basics [Image 2 of 3]’ Photo by Senior Airman Tong Duong via DVIDSHUB