Understanding SafeHash: The SafeHash Pyramid

====================================

Comically enough, SafeHash can be thought of as a pyramid -- not a pyramid scheme, but as three distinct accounts 'mining', 'backing', and 'shares'. The internal management of SafeHash is merely to ensure account values are such that

**mining > backing > shares**

1. Starting out.

When starting the fund, we had 0 mining, 50,000 backing, and 0 share sales. This brought us to the situation 0 > 90,000 > 40000. Now we can buy as many mining as we want so long as the equation holds. So we buy 47,000 mining. This leaves us at 47000 > 43000 > 40000.

2. Continued Operation / Efficient Fund Operation

So long as we have more in mining than backing, and more backing than shares which have been sold, the fund is considered "stable"; good profits and enough to back every share. But in the above example, there are 47,000 mining shares purchased at IPO price and 40,000 shares. We could theoretically sell 7,000 shares and come to a situation of 47,000 > 50,000 > 47,000. Then we would need to buy more mining shares, say 2,000; this would leave us at 49,000 > 48,000 > 47,000. This is more efficient, but still not ideal. How do we measure this efficiency? We note here that the goal seems to be to bring the number of shares sold close to the amount we have in mining. Therefore the equation (mining - shares sold) / (mining account) will give us a number which shows how far away we are from this goal. In the first example. the difference would be 7,000 / 47,000 or 14.9%. Very bad! In the second (adjusted) example the difference is 2,000 / 49,000 or 4.1%. Better but still surprisingly far from ideal.

3. Most Efficient and Safe Operation Guidelines

So long as we first increase the number of mining shares sold such that backing does not fall under 1 NXT/share, we will increase the profitability of the fund to it's maximum. We may then sell a definite number of shares, and by selling them at 1 NXT/share we will not decrease backing beyond transfer fees. Thus with two simple actions it is easy to bring the fund to a very efficient state. Currently, the fund has 145,111 in Mining, 144,316 in backing, and 140,857 shares sold. The efficiency is (145111-140857)/145111 or 2.9%. Our first action then would be to buy mining shares; how many? "backing - shares out" will lower backing as much as possible. In this case it's about 3450 shares of HRLTCGEAR.

We can then note that our mining value has increased to 148,561. Subtracting the current number of shares out (or subtracting the backing, to ensure we don't sell too much) implies we can sell 7,704 shares. Assuming transfer fees of 2 (one per order) we would go from 145,111 in Mining, 144,316 in backing, and 140,857 in shares to 148,571 > 148,568 > 148,561 in just two transactions. For an efficiency rating of 10/148571 or 0.0067%. Now that's super efficient!

4. Stacking the odds in our favor

Being super efficient is great, and allows us to stack the odds in our favor. For one, we don't have to buy 3,450 shares at 1. It's not an emergency, especially if there is no demand for new shares at the moment. Instead, we can place an order for say 3,500 shares at 98% of the ask price, and really get our money's worth. This will enable us to sell more shares in the future and in general make more money for everyone. My guess is, the best time to play around like this is when we have no emergency need to sell shares (i.e. people are not bidding more than 1/share or there are no bids) and we are already very efficient (at least within 1% of ideal).

5. Dividend Policy

Transfer fees are pooled (deducted from the total before it is divided among shareholders), and amounts less than 2 NXT are retained by the fund on your behalf until we can pay you at least 2 NXT (and may be temporarily reinvested).

We can pay dividends at any time; we can pay weekly or monthly. Times and amounts can be random, so long as we keep our promise to pay out 50% of what we take in. Our commitment to this promise, as shown by our actions, will be one way you can evaluate the management of this fund and make a decision whether or not you want to trust us. Either we do it, or we don't, and if we don't, you are of course free to take your money elsewhere.

In general I would expect SafeHash to pay at least 90% of what is owed, at least every two weeks or more frequently. Our first payment will be on the 18th, the next payment is guaranteed before the end of the month. Beyond that I can't say. Please let me know your thoughts, and I will adjust policy accordingly.