A true deflationary currency would be one in which the money supply contracts, so it would have to start at an arbitrary amount and then basically self-destruct (inflation self-destructs by dilution/overabundance). As an example, food would fit this definition because it perishes over time.

Yes, gold is inflationary, as is Bitcoin. The distinction is that Bitcoin approaches an absolute limit whereas discovery of a large supply or mining gold on the moon could continue expansion of its unit base indefinitely.

I think I see what you're trying to get at: gold and Bitcoin inflate less than overall economic expansion grows. So it isn't the currency deflating, there is simply a divergence between the growth rates. In effect, that simulates a deflationary environment - the presentation is the same, but the reasons are different.

This chart illustrates that sequence of events:


A significant change occurs when market forces shift the Bitcoin decimal. Let's take another look:

Annual Potato Yield	>Total Bitcoins	>Value in Bitcoins	>Decimal Shift	Value of 1 Bitcoin
100	1,000	10.000	none to hold 10.0	1.0
1,000	1,000	1.0000	>1 to hold 10.0	1.0
10,000,000	1,000	0.0001	>5 to hold 10.0	1.0

The end result is that anyone holding 1 Bitcoin at the beginning would still have 1 Bitcoin, only now smaller fractions of a Bitcoin are needed to conduct everyday transactions. The currency remains functional in regard to price stability while existing units are not devalued, meaning that savers neither harm the system nor are harmed by it.

In a fiat system, savers are actually the enemy because stockpiling puts strain on price stability and if the stored fiat is ever disbursed in size, it can cause sizeable price disruption. For an example, imagine that a major foreign holder of US debt (bonds and the like) decided to sell; several billion dollars flooding back into the system without an immediate, commensurate balance of trade reaction would be the same as printing that money into existence. Prices of goods and services would be affected within a year, potentially causing further chain reactions that could destabilize the entire system.

What is the universal response to an inflatable money supply that is experiencing excessive demand (i.e. a liquidity crisis)? Inflation is the only answer in the end. By inflating, the decimal point is moved to the left instead of the right. We've seen this with numerous national currencies which introduce a 'new' X fiat currency, just like the old one only several zeros have been lopped off. A 100.00 denomination becomes 1.00 for a left shift of 2 places. That dilutes existing units (savings) in order to maintain price stability.

Using a money supply that is essentially fixed, and is indefinitely divisible, completely negates that problem. Gold is the same, but can only do well to certain point because it is physical, offering practical usage down to about a gram denomination. While it is theoretically possible to use gold held in custodianship to lower the limits on practical usage, that returns to issues of trust regarding financial institutions.

Bitcoin also virtually eliminates the management concerns involved with trust (there will still be weak points, notably the developers, hashing power concentration and cryptographic security). The only real questions that remain (aside from those mentioned) are of eventual widespread adoption and whether the decimal expansion will be sufficiently smoothed by market forces. Therefore, a complementary inflationary currency might not be necessary.

It's hard for those with minimal understanding of their own financial system to grasp these distinctions, and even harder for those that have made it their livelihood and gospel in understanding the existing paradigm. All economic arguments against Bitcoin so far have been bunk. The shift in recognition will be a gradually accelerating process, much like this excellent analysis.

Bitcoin's decimal expansion is effectively inflationary. That inflation is limited by the number of decimals it can be expanded to, which is currently 8. The system itself does not determine the extent of decimal expansion; that results from market forces. As Bitcoin gains wider adoption its value will rise, prices will fall, and the decimal will periodically be moved to the next point.

There's a good chart that I often use to explain the process. Comparing this mechanism to gold, it's like using ounces, half-ounces, quarter-ounces, tenth-ounces and then grams. Being physical, gold is limited to how far it can be divided and remain functional; we can't exchange individual atoms. Being completely abstract, Bitcoin is only limited by the necessary protocol changes that would allow further decimal expansion.

It's possible that market forces will make the change gradual, otherwise a full decimal shift could cause a sudden drop in prices of up to 90% depending on overall economic growth. An inflationary currency definitely affords a smooth level of expansion to match real growth, but may not be necessary. I think we'll have to get much closer to the 21mm unit limit before anything empirical can be determined.

Yes, there's an interesting dynamic there. Ripple provides a credit system, so it and Bitcoin would certainly be complementary. It doesn't work the way a contemporary currency does, though - the mechanism is different and there's no discrete Ripple 'unit'. It acts like a container for quantifiable currencies. That also means they solve somewhat separate problems, and there are different scalability factors.

There is still room left for other variants, including an inflatable system that can maintain stable prices indefinitely. For example, I think that if Bitcoin doesn't implement P2SH this year (the concern over BIP 12/16/17), another system might arise to do so.

I have two of those with shaders that couldn't be unlocked. Flashing allowed for mem clock settings to be run down to 300. Haven't had any issue adjusting voltages with cgminer.

Boom. USD collapse vs. BTC visual, instead of BTC rise vs. USD.

Right now and for some time to come, yes. You'd have to manage a balanced position by scaling it instead of going all in each time.

With Bitcoin, I use weekly data as an overall trending and reversal guide, then daily/hourly to spot scaling opportunities. It'd be more complex using multiple temporal dimensions, but even building on the simple EMA10/21 technique that way would be far more robust - even somewhat predictive.

The greater the time-frame, the harder it is to reverse the trend.

Demurrage is a great concept for traditional fiat-like currencies, but not a Bitcoin-style system: it attempts to ameliorate the issues inherent with human management. Since there is no human component directly involved in management of the money supply of Bitcoin, demurrage is redundant.

Eventually, Triffin's dilemma overcomes demurrage anyway. The divergence between the need for a savings vehicle that is being eroded and the need for price stability which slides further out along a parabolic curve causes too much strain over time, and no amount of effort can prevent a collapse then.

Bitcoin technically solves the dilemma by allowing for theoretically unlimited decimal expansion of the unit, similar to going from ounces of gold to grams. Instead of forcing either perpetually inflation or adamantly static supply, Bitcoin allows for what is effectively an periodic, incremental inflation; these resets are somewhat jarring, but should be mediated so long as market forces can function freely.

Prices would be relatively stable within a range and yet the same unit of savings still exists and remains fully intact, the same as if gold were infinitely divisible and we were able to actually make use of it an atom at a time. Anyone who had a single ounce would be immensely wealthy if a single gold atom were used as a $1 bill is today.

Obviously we can't use individual gold atoms as money in any known practical manner. Likewise, the difficulty that arises with Bitcoin is getting a protocol change that allows for further decimal expansion to actually take hold. A potential solution may be an alternative system that acts as a counterpart by removing the block generation limit, allowing for indefinite expansion. That would work the same as gold and fiat today, only there would again be no direct human management of the unit supply.

If all the gold in the world were hoarded away indefinitely, an alternative would eventually be adopted over a period of time due to gold's disuse. The accounting period could be anything - hours, days, years, etc. Theoretical with gold, neither realistic nor likely.

It's still a potential for Bitcoin, but that's a rapidly shrinking possibility for the monetary aspect of the system. Also, the general structure of Bitcoin is immensely valuable, making the argument impractical outside of academia.

As for the stored value concept, traditional definitions do not satisfactorily encompass Bitcoin. Can value be removed from gold? Why or why not?

Very nice work.

Also try expanding the time frame to daily & weekly scales.

The Buddhist perspective explains Bitcoin's magic perfectly. It's also what makes Bitcoin so hard for most people to grasp. The thing about it is that there is no trust among people, but an agreement between machines; as though something were truly created from nothing. Machines aren't alive and yet a glimmer of it has become visible.

As for MFing Global, I find Jesse's Cafe has provided the greatest level of detail with strong insight. It isn't the first, won't be the last, and has decimated any legitimate functionality in traditional western markets for a long time to come.

Where's the expiration date that erodes the stored value?

Exactly, it's an issue of Bitcoin's stock to flow. That's a process, not an instantaneous event.

For argument, let's say there are 8,000,000 BTC in existence and 8,000 are generated every 2 weeks. The valuation of the 8mm matters more than that of the smaller 8k - if there's demand for the 8k, it won't affect the 8mm very much at all due to the sheer magnitude difference. You can think of it like throwing a rock into a lake - small ripples may occur, but the water level certainly won't be cut in half.

A more likely scenario is that solo miners may begin to drop off, as well as some smaller mining operations. It will gradually become less viable to expend the resources to compete with larger operations that gain benefits from economy of scale. This is also a process that will occur over a period of time, probably while the exchange rate for Bitcoin rises against other currencies.

Don't hope or panic - that doesn't change a trend or highlight tops & bottoms. Read Goomboo's Journal thread. If you're not simply holding long-term positions, he offers some excellent perspective. Besides, anyone who recommends Jesse Livermore is aces.

No worries, Bitcoin isn't going to disappear. While I have my concerns about the BIP 12/16/17 issue regarding P2SH, I think it's unlikely that it will cause real damage to the system during implementation. Other than that, we've seen this week that there is no end in sight to inflationary policies. Long-term that's beautiful for our intangible tokens.

That said, if the $4.64 spike low from the 17th fails to hold, the most likely new low will be around $4. The $3-4 range is where buying should begin to emerge in size. For large mining operations, the profitability zone is still around $2-3 which creates a 'hard' floor as we saw a few months ago.

Besides, there was a solid 2 month run up that needs to be bled off. I wouldn't be surprised to see a spike to just below $3 before the capitulation selling is done (maybe over the coming weekend). At least with Bitcoin, technical analysis is more relevant due to its transparency.

So I'm with StewartJ - bring on the clearance sale!

Ubuntu 11.04 x64 (kernel 2.6.38-13-generic)
Catalyst 11.11
APPSDK 2.5
cgminer 2.1.2

Multiple 69xx cards running under this configuration provides the same results as Cat 11.6 with APPSDK 2.4 and cgminer 2.0.8-2.1.0 only 2.1.2 offers even further reduced rejects/stales. No 100% CPU utilization issue.

Using APPSDK 2.6 appeared to negatively impact overall hashing speed by ~5-20MH/s per GPU with no clear reason. Considered switching to Arch, but time investment doesn't seem worthwhile at the moment with a smoothly running setup (uptime 4 days, no resets since last cgminer git build). No testing done with 5xxx cards.

The limit is strange - pegged at 125Mhz between the core and memory rates, so 900 core won't allow any lower than 725 memory or it'll kick the memory clock back up to stock settings. Earlier, sveetsnelda suggested flashing the GPU BIOS. This works well and I have some 69xx series cards running with 920 core and 300 memory speeds.

It's fairly easy to use a FreeDOS USB stick with AMD/ATI flash tools to boot from. Just pull the current BIOS, open using GPU-Z and then flash the same cards with the updated settings. After that, changing the core speed doesn't automatically cause a reset to high memory clock rates. My card temperatures dropped by at least 10C.

Like I said - minor. I just changed the working directory instead.

Alright, so I'm sticking with the good ol' Ubuntu 11.04 installation for now, but kicked up to Cat 11.11 with APPSDK 2.5 and the latest cgminer 2.1.2 git build.

The good news is that everything is running smoothly so far with <0.5% stales. The bad news is: I don't have a FPGA box to try out yet.

Con, a minor quibble: could the generated .bin be searched for in the config file directory before building it?

That line works for my setup.

Set it to 1% in example.conf file with a note to comment it out if donations aren't desired. I hope you're at least getting a reasonable trickle from those who do have it enabled.