Streamline your deal cycle for businesses with airSlate SignNow

How to create outlook signature

I'm gonna start with just one slide here and we leave this up for a while it's the long term chart of the US equity market going back to 1871 it's a tremendous amount of history we thank Robert Shiller of Yale University for having compiled all this data I want everybody to just kind of take that in for a while while I distract you with a little story anybody seen the movie A Beautiful Mind all right there's got a feel good I saw that with my English daughter as part of her psychology class fascinating movie very interesting guy biographical story of dr. John Nash a Carnegie Institute and Princeton trained mathematician who actually wind up winning a Nobel Prize in Economics for some of his work in game theory among its many compliments he actually worked for the Department of Defense during the war years as a encryption decoder and it was amazing at the people the story and I don't know if the movie was truly accurate from the reviews I've read it may have taken some liberties but during his time at Department he was able to break encryptions the German encryptions just by looking at them which amazed me he could and they depicted it on in the movie very very cool showing you know large walls of data and numbers and you know in his mind you could see the numbers popping out and him deciphering the sequences I thought that was fascinating someone who likes to look at pictures it was really inspiring so going back to you know what I want to start with is if we go back and we really think about starting with a clean slate what has really gone on in the last hundred and forty plus years that we can get this data back well the first thing I see if you look over the entire period is we have basically an upward trend average is about four and a half percent over the entire time period since 1871 but you can't see in many places it's really pretty far off the trend so there's a lot more going on than just descent a simple upward trend so if we go back and again start with this clean slate think about with an open mind what else do we see going on well to me I think you see a very clear distinctive break right around the post-war era and you get two very distinctive trends when you trend fit a line in these periods the more recent post war war to mount an area you get about a seven percent trend which if you talk to you a lot of your financial advisors and you talk about assumptions for equity returns that's usually what you get in a ballpark for US equities is maybe seven percent the pre-war period from from 1871 through about 1942 is only one point eight percent that's a very big difference it's still not quite really happy with what I see though you still see you know some areas where you get very big divergences so we thought about taking a different approach rather than looking for the upward trends think about looking for what are the patterns you might see here that to me that the thing that inspired or intrigued me well jump the gun these sort of periods of flat trends in the equity markets were they they bounced along but really they didn't go anywhere over a long period of time so we just took the I don't know seventeen hundred months worth of data and went and found each time the market had a new high and then went and saw looked for the last time the last date that number occurred so for instance if we look over here the first time it crossed oh I think was about eight and then the last time it was at a was here it never came back to eight so that was the last time so taking that methodology looking for the first time it crosses a level and the last time we come up with these eight of four periods here of basically flat trends very interesting and they vary in time in length the first one there is about 25 years the second one from 1901 to about 1942 is 41 years more recently we've had these two periods in here which lasted about 12 years each so good news for us post-world War two it seems like we're getting fewer and shorter of these flat periods but as we know the future is totally unpredictable so we don't really know what's gonna happen but it's interesting to think about what goes on during these periods and of course in between you have these up periods these upward trend very strong upward trends you're getting returns of low double digits ten twelve thirteen percent in these upward periods intermingled with these long periods of of sort of flat so that becomes really sort of the interesting question that I started to do research on is if you think about what drives the markets in these long trending periods well for Houston invest management if you're going to use any kind of indicators like this they need to be predictable otherwise it really is nothing more than an observation and if you think about what big factors drive the markets over these either long flat periods or long upward trends it'll be things like policies in place say opening up a free trade globally it can be demographics you know certainly we know we can see at least some lead time with birth and death rates length of life getting longer we don't know as much about immigration policies but that certainly can have an impact probably the biggest driver is technological innovation when the economy is structurally changing new technology are coming along and how quickly they're deployed and put to use in the economy so these are all really tough things to sort of predict and again like I said if you can't predict it doesn't really help you much so the question the second question becomes what are the implications of these different periods for investors so in other words if we're we don't know what the future holds what would it mean if we go into another long period of flat equity markets I'm it's conceivable Japan right now had hit its last equity market peak in 1990 and it still has a return to that it's been flat for 20s or at least flat trending or lower lower for 27 years now so it's it's not impossible well if we can't predict these long trends then what else can we look at well you can see during each of these trend periods whether it's a flat trend or an upward trend you've got these cycles so if we separate out these trends take the trends out of the market and just look at the cycles this is kind of what it looks like it's pretty messy I know it's it's it gets a little ugly but you definitely see a lot of ups and downs and of course being part of the American Institute for Economic Research the first thing I think about is what's the impact from business cycles normally you would see the recession shading showing highlighted new recessions all the way back to 1871 they they're highly correlated to these business cycles to the equity price cycles now that takes away of a really sad point you just tell us where the shaded area would be so we could yeah well I can we can try and work you through some of these that's that's really unfortunate well we had recessions we can work over backwards so we know there was a big recession the Great Recession here there was another one back in 2000 2001 we had a 90 recession we had the double-dip reception recessions in 80 81 in fact most of these correlate to an economic business cycle recession so again for any of this information be useful it's got to be predictable well luckily at AI AR we've had business cycle research and a business cycle model in place for many decades there we go this one you can see the recession shading okay so this is our business cycle model we just updated this model it's been in some form it's been around since 1950 it started out life known as statistical indicators we now call it the business cycle conditions model for people who are familiar with it I'll just give you a little side brief in it the business cycle model has three components to it it's got a leading index a Cohen ciders index and a lagging index each of those indexes are made up of a group of indicators there's 12 economic indicators in the leaders six in the coincide errs and six in the laggers the indicators are combined into a diffusion index which is a breadth measured so it's the percentage of the indicators that are either trending higher trending flat or trending lower and we've found that over time it's actually done a pretty good job predicting upcoming business cycles an average about three months ahead of time and the nice thing is with our new business cycle model we are able to calculate it anytime we can do it daily if we want it used to be only monthly and we can distribute it and disseminate it much more quickly now with the internet so this to us becomes a potential tool that we might be able to use in investment management so here we are looking at it versus the S&P index and you can see much better now that we kept the recession shading so our in our indicators start declining well before this is the official recession the gray shading the S&P 500 shows a big decline and you can see it happens in pretty much all of these last we can talk a little bit about the current economy actually I can tell you we updated the model it's now at 42 so we're still kind of hovering below 50 but it's not that bad it's actually to us it's really very consistent with a slow economy not necessarily calling for recession so you can rest at ease I don't think we're going there by mean we can talk more about later that's I think we're safe okay so in order to really test this though pictures are nice but yeah eventually the rubber has to meet the road and we do some back testing so we actually set up a scenario where we can test the indicator statistically to see if they can actually improve performance so here's the way we ran these back tests I'm gonna show you a set of four back tests now the first thing I want you to know that this is all in sample now what that means is the model is calculated today throughout its entire history with all the knowledge we have today so it is not a realistic in other words if you were going to do an out-of-sample it would you could only calculate based on the information you knew at the time we are going to change the methodology so we can do this out-of-sample testing but for right now the easiest way to just see if there was even a proof-of-concept here was to do it in sample and we'll show you the results they're really kind of interesting we're going to use a three-month average of that leaders index we're gonna start with a portfolio of $100,000 in May of 1968 this as far back as our model will allow us to go the initial starting allocation is going to be a hundred percent in equity just the S&P 500 and nothing in cash and throughout the model it will simply go between equity and cash no other instruments no other assets just between equity and cash here's the decision rule again there's lots of ways to implement this but you we found who through a Gaussian technique we've narrowed it down to this the simple rule if the index is below 34 remember we're at 42 so we're okay if we're below 34 and falling we're going to decrease our equity exposure if the index is above 8 and rising we're going to increase the equity exposure and that's it and the other thing we're going to do in four different back tests is change this percentage that we increase and decrease by does everybody understand any questions on that yeah it's just a benchmark index so you it would be done in simple ways through an ETF just to just being in and out of say the spider any other questions okay so let me just then explain a little bit more on this this buy and sell rate that the it's it's a really at how quickly we get in and out of the positions based on the model signals okay for instance if we had a sell rate of 10% the first time we get an a sell signal you'd reduce your equity exposure by 10% so you go from 100 down to 90 if you got a second month the second signal you would then reduce by another 10% 10% of this brings you down to 81 and then so forth so at a 10% sell rate after about six months you'd be about half your equity exposure half in equities and half in cash okay if you increase your buy or sell rate say up to fifteen then you're about two halfway point after four months so you're getting in and out more quickly and the same thing if you go up say up to 33% fairly quickly moving in and out very quickly based on these buy and sell signals you'd hit your halfway point in just two months now keep in mind we're showing here or list over about ten months in the post-world War two error the average economic cycle has only been 11 months so you don't have a long time to move in and out and equity markets move even more quickly than the business cycle so the trade-off here is how quickly you want to move in and out versus how quickly the markets and are moving any questions on sort of the buying so rate and how it works okay so here's our first it's usually votes using cash so whatever cash balance is outstanding as you sell out you would sell a portion of your cash balance so in other words if you're okay so if you now are say 50/50 whatever the dollar value of your cash is you then sell that amount and buy back into equity that make sense okay okay so the results from the first back testing we're going to do these this first one is fairly conservative we're gonna savage sell rate of 10% so again take six months to reduce your accurate exposure in half by rates a little faster one of the things we found and this is a consistent with the results of our economic model the model is better at calling peaks in the cycle and not quite as good not as much lead time in the in the troughs and in general you'd like to stay fully invested early as much as you can so we're going to in all of our back tests we keep the buy rate faster than the sell rate so we go back again next to 68 we have a total of 579 iterations months in them in the back tests we have a total of 228 transactions we do include a transaction fee of $20 figuring after doing it on each rate or whatever it's only 95 we have a total transaction cost over the entire period of about $4,500 distribution of the transactions about 29 30 percent or so or buys 10 percent or sells 60 percent of the time you're on hold so you're not doing that much trading here and again this is all done over 48 year period in the backtest so how did the portfolio deal the reference the basic S&P 500 if you started with $100,000 in 68 you would have ended up with 2.2 million our portfolio moving adjusting the equity allocations you'd have 2.5 2.5 6 so you've gained an extra $350,000 a difference of about 16% over that entire period not too bad and that is net of transaction fees annualized return goes from six six up to seven so not too bad if we take a look at it visually it's kind of interesting as we get some and there in there is no shading on here so we don't have to worry about the missing bars you can see where the model really pays off in some of these big downturns the blue is the benchmark the red is the model we've lowered our equity exposure so our total portfolio doesn't decline as much in both these recessions back here the model works just as we'd hope so we open up the gap here anytime there's a gap widening that means you're outperforming when the gap narrows you're underperforming over most of the period it actually trends the same so we open up a gap we outperform here in the 1980s and the double-dip we lose our advantage it comes right back we're about in line with the portfolio we stay in line in the 1990 recession again the model doesn't work as well there was a fairly mild recession and again the short of the recession is the less time to react to the recent to the movements in the market so it's not that surprising we underperform we underperform by this gap this gap remains relatively consistent but we underperform through the 2000 peak here in this decline again we close the gap the model will close the gap so we're outperforming we underperform just slightly here but then in this latest the Great Recession we outperformed by a significant margin and again this is one of the one of the characteristics of this model so far the way it's been developed is it to really work for it to really work the best you need a fairly substantial recession with a very fairly substantial decline in the markets so it's really protecting you against the big moves it's not designed to do every little up-and-down in the market it's not and it won't in fact capture small moves and it's less effective during the shallow recessions it really is more effective during the bigger more serious recessions and bigger more serious declines here's another way of looking at it apologize for the harsh color it stands out nice and bright so this is your equity allocation if you look at your think about your portfolio over time the percentage of your equity portfolio in equity it's on the right scale so anytime the greens at the top you're fully in equities and these white times are when you're selling out and reducing your equity exposure and increasing cash you see there's one two three four or five basically six episodes throughout this almost 50 year period so again you're not trying to time every little movement this is really focusing on protecting your investment to the equity allocations during a business cycle all right so one of them is if we change our buy and sell rates will increase them a little bit so we'll get in and out of the market a little more quickly we're down to 178 transactions total we've reduced our transaction cost by about a thousand dollars we're now at about 20% buys whole cells didn't change at all about 10% and holes are up to about 70% of the time overall performance stay same 2.2 million in the in the base portfolio we're now up to three almost 3.5 million so that's a difference of one and a quarter million we beat the base portfolio by over 50% we've increased ours the annualized returns by a full hundred basis points over the entire period if we look at it graphically had we do again you'll see the same patterns emerging it really outperforms in these deep ones doesn't do as well in the double-dip you can see the gap closes here close it up here we outperform better now because we're moving in and out of this cycle and we really outperform in this last Great Recession here's the equity allocations you'll notice a couple things going on here the whites at the top are getting narrower so you're in and out more quickly and you're getting steeper cuts in your equity allocation in some cases going below 20% for the equity exposure during these business cycles all right one more time we'll go 33 and 50 so again now we're really getting in and out even more quickly nice result I think transaction fees are going down you're still on hold about 75% of the time and we're now are performing by over a hundred percent over this period annualized growth rate six six versus eight - there we go if we look at the e blessing if we look at the performance of the model portfolios this model portfolio never it gets a gap at the beginning and never looks back it maintains a performance benefit over the entire period and the gap widens particularly as we hit again this last most severe recession and business and equity market decline the equity market exposures again you can see because you're getting out more quickly you've got very narrower white areas here when you're going to cash but much more severe in some cases your equity exposure is getting almost to zero and then finally just to take the extreme case if you were to get in and out a hundred percent on the first signal so a very aggressive trading on this type of behavior but just to see what the potential is we're now down to just twenty eight transactions over a forty eight year period that's two percent buys three percent sells ninety five percent of the time you're sitting and holding and you're going with the market so this is really very concentrated moves anticipating the business cycles if you do this if the this model is accurate we're going to do more testing you would more than triple the performance of the underlying base SP index and just the model outperforms and again you'll notice the key thing that really becomes obvious here how in these periods of down the model portfolios becomes flat to avoid the downs and the rest of the time it sort of follows the the market and that's really where we're trying to make up the most is during the severe market turn down severe business cycles protecting against those downside losses alright let me just emphasize these results are all preliminary okay don't go home and do this don't try this at home results are I think interesting enough that we're going to continue this we've got a lot of things planned we're going to certainly switch our methodology to do the out-of-sample which will be much more realistic we're going to try other methods for moving in and out rather than percentage of the portfolio we may target percentages we may look at using this model in conjunction with other economic indicators one of the very exciting things we're doing is we're now expanding we're building business cycle models for other countries and we will then once we find a business if we can get business cycle models for other countries working we will put that through the same testing we're doing here to see if we can work with other equity markets around the world and again these are very preliminary please don't think you can go home and just follow this model this is all being tested we have a lot of work to do but I think it's really exciting and we're really looking forward to testing this in the future just want to thank Robert Shiller for all his data any questions yes do we didn't do do didn't what we wanted to see if there's a just a proof of concept usefulness to the model everybody's taxes are different we could model them in but at this point we just want to see if it seems like the model is going to be effective or leading indicators also take into account the international economic growth they do sort of implicitly by the nature of the indicators we're using but no we're not really accounting for that sort of factor and it's a very good point because as you know everything change over time the economy is always growing and changing and certainly throughout the last several decades with increased trade and then moving operations overseas yeah a lot of companies are getting a lot of their revenues from from outside sources we hadn't I hadn't thought about that until you brought it up but it's actually another great one of those questions we can look into if there's a way to possibly combined business cycle models for other countries all into one model and to sort of look at a global portfolio but so far there seems to be more I mean the you what the markets around the world are correlated to a degree but I still think most of the equity market performance is driven by domestic fundamentals even with you know foreign exposure growing but it's actually a very good area for us to look into as well we're gonna put me on the spot if we can name them okay well I can tell you listed out of the 12 leaders we do we made a big effort in this in the redesign of the model to keep the indicators broadly representative of the economy so we're not concentrating any one thing there are financial indicators labor market indicators sales and order indicators I can see if I can name them and anybody wants to jump in and feel free there's an initial claim for unemployment insurance the average work week in manufacturing Boise is gonna be tough new waters for every under waters for consumer goods real new orders for core capital goods employment population ratio is that hasn't been 500 we have financial indicators yo curb spread debit balances on margin accounts I know I'm forgetting a few heavy truck unit sale thank you yes heavy truck unit heavy truck sells so tractor trailers buses that sort of thing you know it's your shipping economy everybody's using Amazon it is similar in a way it depends on whether they're shipping by rail or by Road but we found it actually was distinct enough from other capital goods indicators and yet still was a good indicator for the model it's one of the new ones we added I can tell you in the lattice revision we had out of the twelve I think we dropped five and you know five old ones and added five new ones I think it was seven of them stayed the same yeah correct yep it's a new it's completely quantitative again if anybody's interested in the when we did this updater of the business cycle model the way it was done historically was each of the twelve individual indicators and the leaders index was evaluated we got the researchers together everybody evaluated the indicator deciding whether it looked like it was in an upward trend a flat trend or a Down trend and then we combined them into the index for the new model we actually have the computer do all that we've done a significant amount of statistical testing we allow for what we called a fat zero a bandwidth in order to generate a signal but it's all quantitative now and the role of the researchers are now to evaluate the results of the model rather than have input into the model so the model to me is more robust more quantitative but it still allows us to look at and interpret it and actually if I can let me go back I'll flip quickly back to the current conditions so here we are again this is interesting when we were just looking purely at the model we're kind of hovering in this range below 50 historically when it got below 50 that's when we started thinking about is a recession coming I did wouldn't wasn't automatic but that's when we started getting concerned there was still a subjective element to it in most cases you see the model goes when it goes it goes down and we did have this one period here where was hovering for a while before it fell but it was at a significantly lower level so this is kind of interesting the fact that it's sort of unprecedented we're hovering in this slightly below 50 world which which raises flags and we've been saying if you've been reading our pieces we think the risk of recession is elevated but it is not the most likely scenario right now we still think the sort of the slow growth outlook is is most likely and this is actually very consistent with that absolutely at this stage you can't recommend anything other than don't do this okay this is this is not at all ready for primetime I can I can imagine lots of ways of using it you know if you think about most asset allocations now and it's very suitable for most people if you do an asset strategic allocation it's really essentially a buy and hold you're determining what your best portfolio is and you stay with it allowing for some rebalancing you could say hypothetically if you had a 60/40 and out of your sixty percent equity say 20% was foreign and 40% was domestic you could take half of that 40 so maybe 20% and do it by this but leave everything else in the buy and hold the strategic allocation that's one possibility and and how quickly in and how quickly you get in and out again so much more research needs to be done but this is actually to mean really exciting because it can work hand in hand with a typical asset allocation or at some point if we develop enough of these if we have business cycle models that all are adding value globally we could have a separate strategy that does business cycle model investing and it could sit side by side with a strategic allocation that is essentially a buy and hold because again if you think about what can happen in the future we just don't know we don't know if we're going to go through another one of these long periods Japan in one right now we could be anyone right now if this comes way back down we'll find out wow we've really been in a flat period for a long time doing something like a business cycle model and particularly we can take your global would be sort of protection against one of these because you could be this process would generate better returns or theoretically could generate better returns over this long flat period so it's just a way to sort of deal with the uncertainty of the future but again this is all very preliminary we got a lot of testing to go but to me it's very exciting very interesting yeah so I will happily contemplate that when the time comes I have no idea I mean I actually have spent more time thinking about if more and more money keeps flowing into index funds how actually that could impact markets you would think if money is constantly flowing in at a fixed allocation there's less efficiency in the market I mean that's that's just my intuitive question I mean you know if you took it to the extreme if everybody was indexing everything was going into indexing what would help allow you know would provide for price adjustments among the stocks so attention concept there I haven't thought about if everybody were to follow us but I said I'll be happy to think about that if we get to that point anyhow yeah okay so you describe this model as though it's nothing but it really sounds like the family of things where the parameters that you're evaluating have changed over time and some have been preserved and some have been discarded which means that the bottle is changing so does that mean that you have to continue to do the back testing on the latest incarnation of the model or are these results changing as the model changes compared to the S the former so and actually I would say that's exactly what research is designed to do you know there's a big debate about sort of active versus passive management the issue I have with active management is if you've got human intuition in there that's nice but you can't really rely on it to be consistent going forward I much prefer to have some sort of algorithmic systematic whatever you want to call it approach research then can continue to research these different approaches and yes it should be be done because the economy is always changing capital markets are always changing the economy is always changing so yeah right now the model is is in its form the model will not be updated for another year or two at least if the model is updated yes we would start and retest everything to see if it still has the same efficacy as a useful tool in portfolio management but the model right now as it exists is it is sort of in a fixed state and therefore we can run these tests and hopefully find something useful in it just rates are in there in two forms we have a short-term interest rate composite interest rate which is a help me out Co inside or lagger not one of the leaders yogurt is one of the leaders so it's a 10-year one-year spread that is part of our leaders indicators so yes and it's interesting with with what's been going on and interest rate monetary policy the fact that interest rates are so low but as you know again I don't know the future any better than anybody else does fundamentals would suggest at some point if the economy continues to grow there'll be some inflationary pressures that would build Fed will raise rates eventually so interest rates are likely to to go up at some point no idea when but we've got to go with what we have and right now we have the history and and the 10-year one-year spread is is a valuable indicator so it is part of it and this is all part of the business cycle model the economic model which the output of which is used for these back tests yeah this is essentially a market timing I wouldn't categorize it that way one of things I want to emphasize when you looked at how often it traded it's not trading that off and it's intended it's designed to be an economic cycle model so if we have no economic cycles you won't be doing anything you will simply be invested in whatever equity benchmark or portfolio you have this is really meant to to capitalize on economic cycles because we think the relationship is somewhat stable we can go back and test it but now this is now I wouldn't consider this a market timing would I think that's a fairly low frequency trading when you look at the amount of trading that happened in the in the model over that period of time so it's said it's me it's a connected economic cycles not when they're trying to capture every little up and down I don't know what drives the market up and down every single day so wouldn't even begin to try and do something like that any other yeah thank you no no no absolutely not I'm telling you it's not ready for prime time I believe it's not ready for prime time absolutely not please do not do that okay I can't emphasize that enough please don't do that this is really interesting but it is not ready for implementation yes really David I I'm a boy can I really answer that okay my lawyer has my counsel has advised me against answering that any other questions that I can answer all right thank you very much all for coming you