WEBVTT NOTE An Attempt at Reconciliation of Time Preference Theory of Interest and Liquidity Preference Theory of Interest 1 00:00:00.000 --> 00:00:05.000 The first thing I want to point out is that apparently that my full title didn't quite get into the program. 2 00:00:05.000 --> 00:00:10.000 There's a very important word, an attempt at reconciliation. 3 00:00:10.000 --> 00:00:16.000 So it appeared that the conference organizers have more confidence than I do in my ability to actually do this. 4 00:00:16.000 --> 00:00:19.000 But either way, we're going to go forward. 5 00:00:19.000 --> 00:00:25.000 Now how I got inspired to do this? It originally started as a far worse project. 6 00:00:25.000 --> 00:00:29.000 I was planning to really criticize Mises and his exposition of time preference 7 00:00:29.000 --> 00:00:34.500 and side with what I later found Robert Murphy has already done, monetary theory of interest. 8 00:00:34.500 --> 00:00:41.500 But when I read Robert Murphy's dissertation, I found that he made virtually every argument against time preference I was going to make. 9 00:00:41.500 --> 00:00:46.000 And the theory of interest that he laid out was more or less what I was going to lay out. 10 00:00:46.000 --> 00:00:48.000 So I didn't want to be redundant. 11 00:00:48.000 --> 00:00:55.000 And I also found that when he laid out all these objections to time preference, I didn't find them convincing when someone else was saying them. 12 00:00:55.000 --> 00:00:58.000 But at the same time I liked his theory of interest. 13 00:00:58.000 --> 00:01:05.000 So I kind of had this intellectual uneasiness of these two theories of interest kind of juggling around in my head. 14 00:01:05.000 --> 00:01:09.280 So I just kind of lived with it and thought, OK, there's going to be something here, perhaps. 15 00:01:09.280 --> 00:01:16.320 Well, as I went back and I was rereading theorists from either side of this debate, I start finding these striking quotations. 16 00:01:16.320 --> 00:01:25.280 One, when author declares that the costs incurred by holding cash are equal to the amount of interest which the some concerned would have borne when invested. 17 00:01:25.280 --> 00:01:36.280 This feels very much like saying, in effect, the interest rate is the opportunity cost of holding money, which any textbook will tell me is the basis of the liquidity preference theory of interest. 18 00:01:36.280 --> 00:01:47.280 But this quote is found in Ludwig von Mises, which was a bit shocking to me. It actually sounds remarkably like another author that said, interest is the reward for parting with liquidity for a specified period. 19 00:01:47.280 --> 00:01:54.280 This would be John Maynard Keynes. Shockingly! This was not a shocking quotation. 20 00:01:54.280 --> 00:02:01.280 But it is interesting that he points out for a specified period. It's not just giving up liquidity, but it's for some period of time. 21 00:02:01.280 --> 00:02:09.280 Or this other author declared that the increase or decrease in the supply of money in a broad sense can lower or raise the gross market rate of interest, 22 00:02:09.280 --> 00:02:15.280 which we know is a consequence of liquidity preference theory, but Ludwig von Mises again is the one that's saying this. 23 00:02:15.280 --> 00:02:21.280 And another author declaring that as an approximation we can identify the rate of time discounting, 24 00:02:21.280 --> 00:02:24.780 That's the ratio of exchange between present and future goods with the rate of interest. 25 00:02:24.780 --> 00:02:29.780 The author then goes on to describe we need to account for things like changes in purchasing power and what have you, 26 00:02:29.780 --> 00:02:34.780 which sounds remarkably like time preference theory, but this was John Maynard Keynes. 27 00:02:34.780 --> 00:02:39.880 So I started thinking, okay, maybe there is some possibility for reconciliation 28 00:02:39.880 --> 00:02:45.580 if people who are very clearly in one camp or the other start sounding like the other camp at times. 29 00:02:45.580 --> 00:02:49.580 I also found that a lot of the time we actually have similarities in outcomes. 30 00:02:49.580 --> 00:03:19.580 I found this because I teach a money and banking class and I have my students do projects where they have to interpret interest rate movements using different theories and I found that very consistently anytime I had a student that said time preference theory and liquidity preference theory would give different interpretations or different predictions they often got something wrong in either of those theories. That also makes me start doubting how different are they really. This is not original to me. Dennis Robertson made this point in 1937 very shortly after 31 00:03:19.580 --> 00:03:25.140 Keynes developed the liquidity preference theory and claiming that the loanable funds market, 32 00:03:25.140 --> 00:03:30.780 which is what Robertson was pushing, more or less told the same story as liquidity preference 33 00:03:30.780 --> 00:03:33.680 theory just from the other side. 34 00:03:33.680 --> 00:03:37.060 So just as an example, when I think about interest rate movements, say over an Austrian 35 00:03:37.060 --> 00:03:41.940 business cycle, the claim is we have credit expansion pushes down interest rates to an 36 00:03:41.940 --> 00:03:43.180 unsustainably low level. 37 00:03:43.180 --> 00:03:47.000 This results in a boom, which eventually will result in higher interest rates whenever the 38 00:03:47.000 --> 00:03:49.880 Credit Expansion Slows Down Sufficiently. 39 00:03:49.880 --> 00:03:53.600 This sounds remarkably like something Keynes actually said, where he says whilst an increase 40 00:03:53.600 --> 00:03:57.080 in the quantity of money may be expected to reduce the rate of interest, when output is 41 00:03:57.080 --> 00:04:01.400 increased and prices have risen, this feels very much like a boom happening, the effect 42 00:04:01.400 --> 00:04:04.460 of this on liquidity preference will be to increase the quantity of money necessary to 43 00:04:04.460 --> 00:04:10.300 maintain a given rate of interest, which to translate Keynesian into English, would say 44 00:04:10.300 --> 00:04:16.240 the interest rate is going to rise unless we keep increasing the money supply. 45 00:04:16.240 --> 00:04:21.240 So all of this led me to think maybe there's some kind of logical connection between the two. 46 00:04:21.240 --> 00:04:26.240 Originally, I'd hoped to show that they were logically equivalent under a certain set of circumstances. 47 00:04:26.240 --> 00:04:33.240 I couldn't do that, but I could find that liquidity preference is actually a logical result of time preference theory. 48 00:04:33.240 --> 00:04:39.240 Just a hint at that, Rothbard in Man Economy and State has this graph that I have here on the left 49 00:04:39.240 --> 00:04:45.240 where he's showing the individual time market curve, showing how an individual person responds to various pure rates of interest, 50 00:04:45.240 --> 00:05:01.240 Where if the rate of interest is low enough, I end up as a net demander of money, so I'm here, so this is showing the supply of money to the time market that I provide, so it's negative, so I'm a net demander. If the interest rate is high enough, I'm a net supplier to the time market. 51 00:05:01.240 --> 00:05:10.240 To me this felt remarkably like a liquidity preference sort of graph. We have the quantity of money on our horizontal axis, rate of interest on the vertical axis. 52 00:05:10.240 --> 00:05:14.640 Access, really all I have to do to make this a standard liquidity preference graph from 53 00:05:14.640 --> 00:05:20.440 any macroeconomics textbook is take it and spin it around, and there we have our net 54 00:05:20.440 --> 00:05:25.720 demand for money holding, oh that's the wrong one, demand for money holding here, just taking 55 00:05:25.720 --> 00:05:31.040 what was originally here, flip it around, money stock being the vertical point here, 56 00:05:31.040 --> 00:05:35.520 just adding up everybody's quantity of money they're holding, and we can end up with something 57 00:05:35.520 --> 00:05:39.720 that feels very much like liquidity preference, so I guess the question in my mind was could 58 00:05:39.720 --> 00:05:52.720 So how do I build this case in a logical way? Well, it ends up that in terms of the graph, actually Rothbard does it just a matter of pages later when he's discussing the effects of income taxes and what that might do to the interest rate. 59 00:05:52.720 --> 00:05:58.720 So in fact, Rothbard himself does see it's perfectly valid to go from the graph to the left to the one on the right. 60 00:05:58.720 --> 00:06:07.720 So the question is, is there some theoretical underlying thing going on that can explain the similarity between these two theories? 61 00:06:07.720 --> 00:06:12.720 So before I get too deep into that, let's just define exactly what I mean by each of these theories. 62 00:06:12.720 --> 00:06:22.720 Often I find that when we think that, when I first started out and was going to criticize time preference theory, it ended up I was just getting it wrong. 63 00:06:22.720 --> 00:06:29.720 I go back and reread Mises, oh okay, he wasn't actually saying what I thought he was saying, so trying to get this right first. 64 00:06:29.720 --> 00:06:37.720 In its simplest form, the pure time preference theory claims, first, that pure time preference exists, that is, we do prefer want satisfaction sooner to later. 65 00:06:37.720 --> 00:06:47.720 Secondly, that time preference is the source of interest, that originary interest is the ratio of value assigned to nearer want satisfaction to later want satisfaction. 66 00:06:47.720 --> 00:06:58.720 And finally, noting that the market rate of interest isn't a pure originary rate of interest, there are components for things like risk, perhaps purchasing power of money changing. 67 00:06:58.720 --> 00:07:17.720 The liquidity preference theory starts out with the claim that people value money for its liquidity, that interest then would be the reward for parting with that liquidity for some specified period, and finally that it's the quantity of money in conjunction with our preference for liquidity that determines the actual market rate of interest. 68 00:07:17.720 --> 00:07:26.720 So how I end up relating these, I start, let's just say that time preference theory is true, I have an easy enough time saying that now. 69 00:07:26.720 --> 00:07:31.720 We suppose also that there's a monetary economy. This seems rather obviously true in the world we live in. 70 00:07:31.720 --> 00:07:33.720 And we'll also suppose that money is a present good. 71 00:07:33.720 --> 00:07:41.720 This is something that many people who would certainly fall under the time preference theory advocates category would certainly agree to. 72 00:07:41.720 --> 00:07:46.720 Puerto de Soto is very direct about this. Rothbard very directly states that money is a present good. 73 00:07:46.720 --> 00:07:54.720 Mises in Theory of Money and Credit almost kind of offhand assumes that money is a present good without necessarily coming out and saying it, 74 00:07:54.720 --> 00:08:05.680 In that case, I think we would agree that the holding of money is valued because money 75 00:08:05.680 --> 00:08:19.400 is liquid, it is marketable. At the same time, giving up present money to receive future 76 00:08:19.400 --> 00:08:24.280 money should result in a payment of interest. I am, after all, giving up present want satisfaction, 77 00:08:24.280 --> 00:08:30.340 want for this marketable liquid currency in order to get that liquidity later. 78 00:08:30.340 --> 00:08:35.200 I am not after all giving up liquidity itself, I'm just giving it up now to get it in the 79 00:08:35.200 --> 00:08:36.200 future. 80 00:08:36.200 --> 00:08:39.740 But in that case it's not totally crazy to say that interest is in fact a reward for 81 00:08:39.740 --> 00:08:42.760 parting with liquidity for some specified period. 82 00:08:42.760 --> 00:08:48.640 It's not insane, perhaps imprecise, but it's not nuts to talk this way. 83 00:08:48.640 --> 00:08:54.640 And then I look at what are the effects that we get from changes in this money market. 84 00:08:54.640 --> 00:09:02.640 If we look at an increase in the supply of money, if money is a present good, that's really just an increase in the supply of present goods versus future goods. 85 00:09:02.640 --> 00:09:09.640 But if there's an increase in the supply of present goods, then we know from diminishing marginal utility that the marginal value of present goods will fall. 86 00:09:09.640 --> 00:09:16.640 If future goods have not changed, then we should see a decrease in interest rates as a result of that. 87 00:09:16.640 --> 00:09:32.640 Then we look at the demand side. If we have an increased demand for money, it's really just an increased demand for present goods, and time preference theory tells us that if we have a greater demand for present goods versus future goods, time preference has increased, and interest rates will rise as well. 88 00:09:32.640 --> 00:09:40.640 All of this sounding very much like what liquidity preference theory would tell us, but ultimately coming from time preference theory. 89 00:09:40.640 --> 00:09:51.640 So then there's the question, why is it that these two theories of interest are treated as two theories of interest rather than, as I would claim, liquidity preference being a special case of time preference? 90 00:09:51.640 --> 00:09:58.640 On one point is that I think on Rothbard, it shows a certain amount of confusion in how we treat money in regards to time. 91 00:09:58.640 --> 00:10:06.640 So on the one hand, he says money is a present good. So for example, the expenditure of consumers on consumers' goods is not a time transaction. 92 00:10:06.640 --> 00:10:10.200 It's just an exchange of present goods for present goods. Very, very clearly 93 00:10:10.200 --> 00:10:14.680 specifying consumers goods are present goods just as money is a present good. 94 00:10:14.680 --> 00:10:19.240 But then he goes on to say that a greater proportion of funds hoarded can 95 00:10:19.240 --> 00:10:22.360 be drawn from funds that formerly went to consumption, which will bring about a 96 00:10:22.360 --> 00:10:27.800 fall in the rate of interest. But if in reality both money and consumers goods 97 00:10:27.800 --> 00:10:32.320 are present goods, then deciding to spend a little bit less on consumers goods and 98 00:10:32.320 --> 00:10:35.760 save it as money is not a time transaction. There should be no effect 99 00:10:35.760 --> 00:10:39.760 Right here on time preference, it doesn't reflect any change in time preference, which 100 00:10:39.760 --> 00:10:44.160 means there shouldn't be any change in the period of interest. 101 00:10:44.160 --> 00:10:46.960 So on the one hand, Rothbard treats money as a present good. 102 00:10:46.960 --> 00:10:50.120 On the other hand, he treats it as neither present nor future. 103 00:10:50.120 --> 00:10:53.280 Guido Holzman also points this out, saying in this case he's treating it more or less 104 00:10:53.280 --> 00:10:57.280 as time neutral, as a time neutral good. 105 00:10:57.280 --> 00:11:00.880 Now Keynes on the other hand is somewhat worse, at least Rothbard I would say is right about 106 00:11:00.880 --> 00:11:04.440 half the time, then he forgets things as he's writing further on. 107 00:11:04.440 --> 00:11:11.440 He ends up misdiagnosing his theory as something new, when in reality it's just a very special case of time preference. 108 00:11:11.440 --> 00:11:13.440 He's thinking about time and time preference, he's only thinking about consumption. 109 00:11:13.440 --> 00:11:19.440 And since money that I have in my pocket isn't being consumed, it must be a provision for future goods. 110 00:11:19.440 --> 00:11:23.440 Since he's thinking about, in a sense, consumption rather than want satisfaction, 111 00:11:23.440 --> 00:11:27.440 realizing that I do actually want to have liquidity right now, that is a present good. 112 00:11:27.440 --> 00:11:31.440 As a result, he ends up misdiagnosing his theory as something new, 113 00:11:31.440 --> 00:11:36.000 New, when in reality it's just a very special case of time preference. If you really get 114 00:11:36.000 --> 00:11:41.120 into the details of it, he makes it very clear that when he lays out time preference theories, 115 00:11:41.120 --> 00:11:45.840 assuming that spending on consumption is totally fixed, so it's really just this trade-off 116 00:11:45.840 --> 00:11:52.240 between money versus investment, which, okay, if we allow him that consumption being fixed, 117 00:11:52.240 --> 00:11:55.920 that's really just making the decision between present and future goods being that choice 118 00:11:55.920 --> 00:11:57.920 The Theory of Money and Credit 119 00:12:25.920 --> 00:12:52.920 In conclusion, if we take pure time preference, add in this monetary system, we end up with something that looks very much like liquidity preference, realizing that money is actually a present good. 120 00:12:52.920 --> 00:12:57.520 And I think that recognition is something that helps clear up some of the confusion 121 00:12:57.520 --> 00:13:01.520 and why we see these two seemingly different theories being treated as different. 122 00:13:01.520 --> 00:13:06.520 If we consistently treat money as a present good, that difference, I claim, would vanish. 123 00:13:06.520 --> 00:13:12.020 One last significant remaining question is how we relate all this back to the structure of production. 124 00:13:12.020 --> 00:13:17.320 There, actually, just this week, I ran across Sceeto Holtzman's writing regarding the demand for money 125 00:13:17.320 --> 00:13:21.720 and the time structure of production, and I think he's moving the right direction 126 00:13:21.720 --> 00:13:31.720 some revision perhaps needed. So at this point I will satisfy your time preference for me ending sooner rather than later and thank you for your time.